❌

Normal view

There are new articles available, click to refresh the page.
Yesterday — 5 December 2025CIO

Resops: Turning AI disruption into business momentum

5 December 2025 at 12:23

The world has changed — artificial intelligence (AI) is reshaping business faster than most can adapt


The rise of large language models and agentic AI has created unprecedented scale, speed, and complexity. Enterprises are moving from static infrastructures to hyperplexed, distributed, and autonomous systems. Organizations are pouring more than $400 billion into AI infrastructure, a wave expected to generate more than $2 trillion in new value. But without resilience at the core, that value remains at risk.

As innovation accelerates, new risks emerge just as quickly. Security is lagging behind transformation. Data is exploding, with nearly 40% year-over-year growth across hybrid and multicloud environments. Regulations are tightening, and ransomware and AI-powered attacks are multiplying. The result: Resilience now defines competitive advantage.

Resilience drives velocity

Resilience isn’t just recovery. It’s also the foundation of sustained innovation. Traditional recovery models were built for yesterday’s outages, not today’s AI-driven disruptions, which unfold in milliseconds. In this world, recovery is table stakes. True resilience means that every system runs on clean, verifiable data, and it restores trust when it’s tested.

The most resilient organizations are also the fastest movers. They adopt emerging technologies with confidence, recover with speed and integrity, and innovate at scale. Resilience has evolved from a safety net to the engine of enterprise speed and scalability.

Introducing resops, the model for next-generation resilience

Resops, short for resilience operations, is an operating model that unifies data protection, cyber recovery, and governance into a single intelligent system. It creates an ongoing loop that monitors, validates, and protects data across hybrid and multicloud environments, enabling organizations to detect risks early and recover with confidence.

By integrating resilience into every layer of operations, resops transforms it from an isolated function into a proactive discipline — one that keeps businesses secure, compliant, and ready to adapt in the AI era.

To learn more about ResOps, read “ResOps: The future of resilient business in the era of AI.” 


Vertical AI development agents are the future of enterprise integrations

5 December 2025 at 10:58

Enterprise Application Integration (EAI) and modern iPaaS platforms have become two of the most strategically important – and resource-constrained – functions inside today’s enterprises. As organizations scale SaaS adoption, modernize core systems, and automate cross-functional workflows, integration teams face mounting pressure to deliver faster while upholding strict architectural, data quality, and governance standards.

AI has entered this environment with the promise of acceleration. But CIOs are discovering a critical truth:

Not all AI is built for the complexity of enterprise integrations – whether in traditional EAI stacks or modern iPaaS environments.

Generic coding assistants such as Cursor or Claude Code can boost individual productivity, but they struggle with the pattern-heavy, compliance-driven reality of integration engineering. What looks impressive in a demo often breaks down under real-world EAI/iPaaS conditions.

This widening gap has led to the rise of a new category: Vertical AI Development Agents – domain-trained agents purpose-built for integration and middleware development. Companies like CurieTech AI are demonstrating that specialized agents deliver not just speed, but materially higher accuracy, higher-quality outputs, and far better governance than general-purpose tools.

For CIOs running mission-critical integration programs, that difference directly affects reliability, delivery velocity, and ROI.

Why EAI and iPaaS integrations are not a “Generic Coding” problem

Integrations—whether built on legacy middleware or modern iPaaS platforms – operate within a rigid architectural framework:

  • multi-step orchestration, sequencing, and idempotency
  • canonical data transformations and enrichment
  • platform-specific connectors and APIs
  • standardized error-handling frameworks
  • auditability and enterprise logging conventions
  • governance and compliance embedded at every step

Generic coding models are not trained on this domain structure. They often produce code that looks correct, yet subtly breaks sequencing rules, omits required error handling, mishandles transformations, or violates enterprise logging and naming standards.

Vertical agents, by contrast, are trained specifically to understand flow logic, mappings, middleware orchestration, and integration patterns – across both EAI and iPaaS architectures. They don’t just generate code – they reason in the same structures architects and ICC teams use to design integrations.

This domain grounding is the critical distinction.

The hidden drag: Context latency, expensive context managers, and prompt fatigue

Teams experimenting with generic AI encounter three consistent frictions:

Context Latency

Generic models cannot retain complex platform context across prompts. Developers must repeatedly restate platform rules, logging standards, retry logic, authentication patterns, and canonical schemas.

Developers become “expensive context managers”

A seemingly simple instruction—“Transform XML to JSON and publish to Kafka”—
quickly devolves into a series of corrective prompts:

  • “Use the enterprise logging format.”
  • “Add retries with exponential backoff.”
  • “Fix the transformation rules.”
  • “Apply the standardized error-handling pattern.”

Developers end up managing the model instead of building the solution.

Prompt fatigue

The cycle of re-prompting, patching, and enforcing architectural rules consumes time and erodes confidence in outputs.

This is why generic tools rarely achieve the promised acceleration in integration environments.

Benchmarks show vertical agents are about twice as accurate

CurieTech AI recently published comparative benchmarks evaluating its vertical integration agents against leading generic tools, including Claude Code.
The tests covered real-world tasks:

  • generating complete, multi-step integration flows
  • building cross-system data transformations
  • producing platform-aligned retries and error chains
  • implementing enterprise-standard logging
  • converting business requirements into executable integration logic

The results were clear: generic tools performed at roughly half the accuracy of vertical agents.

Generic outputs often looked plausible but contained structural errors or governance violations that would cause failures in QA or production. Vertical agents produced platform-aligned, fully structured workflows on the first pass.

For integration engineering – where errors cascade – this accuracy gap directly impacts delivery predictability and long-term quality.

The vertical agent advantage: Single-shot solutioning

The defining capability of vertical agents is single-shot task execution.

Generic tools force stepwise prompting and correction. But vertical agents—because they understand patterns, sequencing, and governance—can take a requirement like:

“Create an idempotent order-sync flow from NetSuite to SAP S/4HANA with canonical transformations, retries, and enterprise logging.”


and return:

  • the flow
  • transformations
  • error handling
  • retries
  • logging
  • and test scaffolding

in one coherent output.

This shift – from instruction-oriented prompting to goal-oriented prompting—removes context latency and prompt fatigue while drastically reducing the need for developer oversight.

Built-in governance: The most underrated benefit

Integrations live and die by adherence to standards. Vertical agents embed those standards directly into generation:

  • naming and folder conventions
  • canonical data models
  • PII masking and sensitive-data controls
  • logging fields and formats
  • retry and exception handling patterns
  • platform-specific best practices

Generic models cannot consistently maintain these rules across prompts or projects.

Vertical agents enforce them automatically, which leads to higher-quality integrations with far fewer QA defects and production issues.

The real ROI: Quality, consistency, predictability

Organizations adopting vertical agents report three consistent benefits:

1. Higher-Quality Integrations

Outputs follow correct patterns and platform rules—reducing defects and architectural drift.

2. Greater Consistency Across Teams

Standardized logic and structures eliminate developer-to-developer variability.

3. More Predictable Delivery Timelines

Less rework means smoother pipelines and faster delivery.

A recent enterprise using CurieTech AI summarized the impact succinctly:

“For MuleSoft users, generic AI tools won’t cut it. But with domain-specific agents, the ROI is clear. Just start.”

For CIOs, these outcomes translate to increased throughput and higher trust in integration delivery.

Preparing for the agentic future

The industry is already moving beyond single responses toward agentic orchestration, where AI systems coordinate requirements gathering, design, mapping, development, testing, documentation, and deployment.

Vertical agents—because they understand multi-step integration workflows—are uniquely suited to lead this transition.

Generic coding agents lack the domain grounding to maintain coherence across these interconnected phases.

The bottom line

Generic coding assistants provide breadth, but vertical AI development agents deliver the depth, structure, and governance enterprise integrations require.

Vertical agents elevate both EAI and iPaaS programs by offering:

  • significantly higher accuracy
  • higher-quality, production-ready outputs
  • built-in governance and compliance
  • consistent logic and transformations
  • predictable delivery cycles

As integration workloads expand and become more central to digital transformation, organizations that adopt vertical AI agents early will deliver faster, with higher accuracy, and with far greater confidence.

In enterprise integrations, specialization isn’t optional—it is the foundation of the next decade of reliability and scale.

Learn more about CurieTech AI here.

Agile isn’t just for software. It’s a powerful way to lead

5 December 2025 at 09:12

In times of disruption, Agile leadership can help CIOs make better, faster decisions — and guide their teams to execute with speed and discipline.

When the first case of COVID hit my home city, it was only two weeks after I’d become president of The Persimmon Group. For more than a decade, I’d coached leaders, teams and PMOs to execute their strategy with speed and discipline.

But now — in a top job for the first time — I was reeling.

Every plan we had in motion — strategic goals, project schedules, hiring decisions — was suddenly irrelevant. Clients froze budgets. Team members scrambled to set up remote work for the first time, many while balancing small children and shared spaces.

Within days, we were facing a dozen high-stakes questions about our business, all with incomplete information. Each answer carried massive operational and cultural implications.

We couldn’t just make the right call. We had to make it fast. And often, we were choosing between a bunch of bad options.

From crisis to cadence

At first, we tried to lead the way we always had: gather the facts, debate the trade-offs and pick the best path forward. But in a landscape that changed daily, that rhythm broke down fast.

The information we needed didn’t exist yet. The more we waited for certainty — or gamed out endless hypotheticals — the slower and more reactive we became.

And then something clicked. What if the same principles that helped software teams move quickly and learn in real time could help lead us through uncertainty?

So we started experimenting.

We shortened our time horizons. Made smaller bets. Created fast feedback loops. We became almost uncomfortably transparent, involving the team directly in critical decisions that affected them and their work.

In the months that followed, those experiments became the backbone of how we led through uncertainty — and how we continue to lead today.

An operating system for change

What emerged wasn’t a formal framework. It was a set of small, deliberate habits that brought the same rhythm and focus to leadership that Agile brings to delivery.

Here’s what that looked like in practice:

Develop a ‘fast frame’ to focus decisions

In the first few months of the pandemic, our leadership meetings were a tangle of what-ifs. What if we lost 20% of planned revenue this year? What if we lost 40%? Would we do layoffs? Furloughs? Salary cuts? And when would we do them — preemptively or reactively?

We were so busy living in multiple possible futures that it was difficult to move forward with purpose. To break out of overthinking mode, we built a lightweight framework we now call our fast frame. It centered on five questions:

  1. What do we know for sure?
  2. What can we find out quickly?
  3. What is unknowable right now?
  4. What’s the risk of deciding today?
  5. What’s the risk of not deciding today?

The fast frame forced us to separate facts from conjecture. It also helped us to get our timing right. When did we need to move fast, even with imperfect information? When could we afford to slow down and get more data points?

The fast frame helped us slash decision latency by 20% to 30%.

It kept us moving when the urge was to stall and it gave us language to talk about uncertainty without letting it rule the room.

Build plans around small, fast experiments

After using our fast frame for a while, we realized something: Our decisions were too big.

In an environment changing by the day, Big Permanent Decisions were impractical — and a massive time sink. Every hour we spent debating a Big Permanent Decision was an hour we weren’t learning something important.

So we replaced them with For-Now Decisions — temporary postures designed to move us forward, fast, while we learned what was real.

Each For-Now Decision had four parts:

  1. The decision itself — the action we’d take based on what we knew at that moment.
  2. A trigger for when to revisit it — either time-based (two weeks from now) or event-based (if a client delays a project).
  3. A few learning targets — what we hoped to discover before the next checkpoint.
  4. An agility signal — how we communicated the decision to the team. We’d say, “This is our posture for now, but we may change course if X. We’ll need your help watching for Y as we learn more.”

By framing decisions this way, we removed the pressure to be right. The goal wasn’t to predict the future but to learn from it faster. By abandoning bad ideas early, we saved 300 to 400 hours a year.

Increase cadence and transparency of communication

In those early weeks, we learned that the only thing more dangerous than a bad decision was a silent one. When information moves slower than events, people fill the gaps with assumptions.

So we made communication faster — and flatter. Every morning, our 20-person team met virtually for a 20-minute standup. The format was simple but consistent:

  • Executive push. We shared what the leadership team was working on, what decisions had been made and what input we needed next.
  • Team pull. Anyone could ask questions, raise issues or surface what they were hearing from clients.
  • Needs and lessons. We ended with what people needed to stay productive and what we were learning that others could benefit from.

The goal wasn’t to broadcast information from the top — or make all our decisions democratically. It was to create a shared operating picture. The standup became a heartbeat for the company, keeping everyone synchronized as conditions changed.

Transparency replaced certainty. Even when we didn’t have all the answers, people knew how decisions were being made and what we were watching next. That openness built confidence faster than pretending we had it all figured out.

That transparency paid off.

While many small consulting firms folded in the first 18 months of the pandemic, Agile leadership helped us double revenue in 24 months.

We stayed fully staffed — no layoffs, no pay cuts beyond the executive team. And the small bets we made during the pandemic helped rapidly expand our client base across new industries and international geographies.

Develop precise language to keep the team aligned

As we increased the speed of communication, we discovered something else: agility requires precision. When everything is moving fast, even small misunderstandings can send people sprinting in different directions.

We started tightening our language. Instead of broad discussions about what needed to get done, we’d ask, “What part of this can we get done by Friday?” That forced us to think in smaller delivery windows, sustain momentum and get specific about what “done” looked like.

We also learned to clarify between two operating modes: planning versus doing. Before leaving a meeting where a direction was discussed, we’d confirm our status:

  • Phase 1 meant we were still exploring, shaping and validating and would need at least one more meeting before implementing anything.
  • Phase 2 meant we were ready to execute.

That small distinction saved us hours of confusion, especially in cross-functional work.

Precise language gave us speed. It eliminated assumptions and kept everyone on the same page about where we were in the process. The more we reduced ambiguity, the faster — and calmer — the team moved.

Protect momentum by insisting on rest

Agility isn’t about moving faster forever — it’s about knowing when to slow down. During the first months of the pandemic, that lesson was easy to forget. Everything felt urgent and everyone felt responsible.

In software, a core idea behind Agile sprints is maintaining a sustainable pace of work. A predictable, consistent level of effort that teams can plan around is far more effective than the heroics often needed in waterfall projects to hit a deadline.

Agile was designed to be human-centered, protecting the well-being and happiness of the team so that performance can remain optimal. We tried to lead the same way.

After the first few frenetic months, I capped my own workday at nine hours. That boundary forced me to get honest about what could actually be done in the time I had — and prioritize ruthlessly. It also set a tone for the team. We adjusted scopes, redistributed work and held one another accountable for disconnecting at day’s end.

The expectation wasn’t endless effort — it was sustainable effort. That discipline kept burnout low and creativity high, even during our most demanding seasons. The consistency of our rest became as important as the intensity of our work. It gave us a rhythm we could trust — one that protected our momentum long after the crisis passed.

Readiness is the new stability

Now that the pandemic has passed, disruption has simply changed shape — AI, market volatility, new business models and the constant redefinition of “normal.” What hasn’t changed is the need for leaders who can act with speed and discipline at the same time.

For CIOs, that tension is sharper than ever. Technology leaders are being asked to deliver transformation at pace — without burning out their people or breaking what already works. The pressures that once felt exceptional have become everyday leadership conditions.

But you don’t have to be a Scrum shop or launch an enterprise Agile transformation to lead with agility. Agility is a mindset, not a method. To put the mindset into practice, focus on:

  • Shorter planning horizons
  • Faster, smaller decisions
  • Radical transparency
  • Language that brings alignment and calm
  • Boundaries that protect the energy of the team

These are the foundations of sustainable speed.

We built those practices in crisis, but they’ve become our default operating system in calmer times. They remind me that agility isn’t a reaction to change — it’s a readiness for it. And in a world where change never stops, that readiness may be a leader’s most reliable source of stability.

This article is published as part of the Foundry Expert Contributor Network.
Want to join?

LLM゚ヌゞェント時代のプロダクトマネゞメント──仕様は“振る舞い”から蚭蚈せよ

5 December 2025 at 07:19

機胜志向から「振る舞い志向」ぞのパラダむムシフト

埓来の゜フトりェア開発においお、仕様ずは機胜ず画面の䞀芧であるこずが倚くありたした。どのボタンを抌すずどのAPIが呌ばれ、どのデヌタがどのように曎新されるかを、フロヌチャヌトや画面遷移図で蚘述するやり方です。このアプロヌチは、入力ず出力が厳密に定矩できる決定論的なシステムに察しおは非垞に有効でした。

ずころが、LLM゚ヌゞェントは本質的に確率的なシステムです。同じ質問をしおも、生成される文章は毎回少しず぀異なりたすし、状況の倉化やメモリの内容、倖郚ツヌルからのレスポンスによっおも振る舞いが倉わりたす。このようなシステムに察しお「すべおの入力パタヌンず出力を網矅する仕様曞」を曞こうずするず、すぐに砎綻しおしたいたす。結果ずしお、「なんずなく賢いアシスタントを入れおみたが、どうなっおいれば成功なのか分からない」ずいう状態に陥りがちです。

そこで必芁になるのが、機胜ベヌスではなく振る舞いベヌスの仕様蚭蚈です。重芁なのは「この゚ヌゞェントはどんな人栌・圹割を持ち、ナヌザヌから芋おどのように振る舞っおほしいのか」を蚀語化するこずです。専門甚語をどの皋床䜿うのか、どこたで螏み蟌んだ提案をしおよいのか、分からないずきに黙り蟌むのではなくどう質問し返すのか、ずいった察話䞊の振る舞いに加え、どの倖郚ツヌルをどの状況で䜿っおよいのか、どこから先は必ず人間の承認を挟むのかずいった、暩限や責任に関するルヌルも仕様の䞀郚になりたす。

プロダクトマネヌゞャヌは、これらを自然蚀語で蚘述された「行動指針」ずしお定矩し、それをプロンプトやシステムメッセヌゞ、ポリシヌファむルずしお実装チヌムず共有しおいく必芁がありたす。埓来の芁件定矩曞に、人栌蚭蚈や察話ポリシヌ、ツヌル利甚ルヌルずいった新しい章が远加されるむメヌゞです。

仕様曞ずしおのプロンプトずポリシヌ蚭蚈

LLM゚ヌゞェントにおいお、プロンプトは単なる「その堎しのぎの魔法の呪文」ではなく、仕様曞そのものに近い圹割を果たしたす。ずくにシステムプロンプトやロヌル定矩、ツヌルの説明文などには、プロダクトマネヌゞャヌが考え抜いた行動ポリシヌが反映されるべきです。

たずえば、カスタマヌサポヌト向けの゚ヌゞェントを蚭蚈する堎合、「顧客の感情を先に受け止める」「自瀟に非がある可胜性を軜々しく認めないが、決しお責任転嫁もしない」「法的な刀断を䌎う衚珟は必ず保留し、人間の担圓者に゚スカレヌションする」ずいったニュアンスをプロンプトに埋め蟌むこずができたす。ここで有効なのは、抜象的な矎蟞麗句ではなく、実際にあり埗る䌚話䟋を含めた具䜓的な指瀺です。良い応答䟋ず悪い応答䟋を䞊べ、どちらを目指すかを明瀺するこずで、モデルの振る舞いは倧きく倉わりたす。

さらに、ツヌル利甚ポリシヌも仕様ずしお明文化する必芁がありたす。どのツヌルは読み取り専甚なのか、どのAPIを呌ぶ際には必ずナヌザヌに確認を求めるのか、連続しお倖郚サヌビスを叩きすぎないためのレヌト制限はどう蚭蚈するのかずいった点を、プロダクトマネヌゞャヌがビゞネス偎・セキュリティ偎の利害を調敎しながら決めおいきたす。その結果は、゚ヌゞェントのランタむム蚭定ずプロンプト䞡方に反映されたす。

このように、プロンプトずポリシヌは「コヌドではない仕様」でありながら、システムの振る舞いを匷く芏定したす。したがっお、プロンプトの改蚂は仕様倉曎そのものであり、倉曎管理やレビュヌのプロセスが必芁です。誰がどの目的でプロンプトを曎新し、それによっおどの指暙がどのように倉化したのかを蚘録しおおくこずは、品質ずガバナンスの䞡面から重芁になっおいきたす。

評䟡・ロヌルアりト・組織䜓制の再蚭蚈

振る舞いベヌスの仕様を蚭蚈できたずしおも、それが「良いかどうか」をどう評䟡するかずいう問題が残りたす。LLM゚ヌゞェントでは、䞀件䞀件の応答の正しさだけでなく、タスク党䜓ずしおの成功率、ナヌザヌが節玄できた時間、誀動䜜によるリスクの頻床ず重倧性など、耇数の指暙を組み合わせお刀断する必芁がありたす。

実務䞊は、たず限定されたナヌスケヌスを察象に、パむロットナヌザヌを盞手にベヌタ運甚を行うのが珟実的です。その際、ナヌザヌにはなるべくそのたたのログを残しおもらい、どの堎面で゚ヌゞェントが圹に立ち、どの堎面でむラッずさせられたのかを定性的・定量的に分析したす。プロダクトマネヌゞャヌは、その結果をもずに、プロンプトやツヌル構成、むンタヌフェヌスを繰り返し調敎しおいきたす。評䟡指暙ずしおは、タスク完了たでに必芁なステップ数の枛少、手動察応ぞの゚スカレヌション率、ナヌザヌの䞻芳的満足床などが䜿われるこずが倚くなりたす。

ロヌルアりトの戊略も、埓来の機胜リリヌスずは少し異なりたす。LLM゚ヌゞェントは、暩限の範囲によっおリスクが倧きく倉わるため、最初は「提案のみ」「ドラフトのみ」ずいった控えめなモヌドで導入し、䞀定の実瞟が確認できおから「自動実行」の範囲を広げおいく段階的なアプロヌチが望たしいでしょう。その過皋で、ナヌザヌ教育や利甚ポリシヌの明文化も䞊行しお進める必芁がありたす。

最埌に、組織䜓制に぀いおも觊れおおく必芁がありたす。LLM゚ヌゞェントのプロダクトには、モデルのチュヌニングやプロンプト蚭蚈に詳しいメンバヌ、ドメむン知識を持぀業務偎のメンバヌ、セキュリティ・法務の芳点からリスクを芋られるメンバヌなど、倚様な専門性が求められたす。プロダクトマネヌゞャヌは、その橋枡し圹ずしお、技術ずビゞネスずガバナンスを統合する「翻蚳者」のような存圚になりたす。この新しい圹割を自芚し、孊び続けるこずが、LLM゚ヌゞェント時代のPMに求められる最倧の資質だず蚀えるでしょう。

Agents-as-a-service are poised to rewire the software industry and corporate structures

5 December 2025 at 05:00

This was the year of AI agents. Chatbots that simply answered questions are now evolving into autonomous agents that can carry out tasks on a user’s behalf, so enterprises continue to invest in agentic platforms as transformation evolves. Software vendors are investing in it as fast as they can, too.

According to a National Research Group survey of more than 3,000 senior leaders, more than half of executives say their organization is already using AI agents. Of the companies that spend no less than half their AI budget on AI agents, 88% say they’re already seeing ROI on at least one use case, with top areas being customer service and experience, marketing, cybersecurity, and software development.

On the software provider side, Gartner predicts 40% of enterprise software applications in 2026 will include agentic AI, up from less than 5% today. And agentic AI could drive approximately 30% of enterprise application software revenue by 2035, surpassing $450 billion, up from 2% in 2025. In fact, business users might not have to interact directly with the business applications at all since AI agent ecosystems will carry out user instructions across multiple applications and business functions. At that point, a third of user experiences will shift from native applications to agentic front ends, Gartner predicts.

It’s already starting. Most enterprise applications will have embedded assistants, a precursor to agentic AI, by the end of this year, adds Gartner.

IDC has similar predictions. By 2028, 45% of IT product and service interactions will use agents as the primary interface, the firm says. That’ll change not just how companies work, but how CIOs work as well.

Agents as employees

At financial services provider OneDigital, chief product officer Vinay Gidwaney is already working with AI agents, almost as if they were people.

“We decided to call them AI coworkers, and we set up an AI staffing team co-owned between my technology team and our chief people officer and her HR team,” he says. “That team is responsible for hiring AI coworkers and bringing them into the organization.” You heard that right: “hiring.”

The first step is to sit down with the business leader and write a job description, which is fed to the AI agent, and then it becomes known as an intern.

“We have a lot of interns we’re testing at the company,” says Gidwaney. “If they pass, they get promoted to apprentices and we give them our best practices, guardrails, a personality, and human supervisors responsible for training them, auditing what they do, and writing improvement plans.”

The next promotion is to a full-time coworker, and it becomes available to be used by anyone at the company.

“Anyone at our company can go on the corporate intranet, read the skill sets, and get ice breakers if they don’t know how to start,” he says. “You can pick a coworker off the shelf and start chatting with them.”

For example, there’s Ben, a benefits expert who’s trained on everything having to do with employee benefits.

“We have our employee benefits consultants sitting with clients every day,” Gidwaney says. “Ben will take all the information and help the consultants strategize how to lower costs, and how to negotiate with carriers. He’s the consultants’ thought partner.”

There are similar AI coworkers working on retirement planning, and on property and casualty as well. These were built in-house because they’re core to the company’s business. But there are also external AI agents who can provide additional functionality in specialized yet less core areas, like legal or marketing content creation. In software development, OneDigital uses third-party AI agents as coding assistants.

When choosing whether to sign up for these agents, Gidwaney says he doesn’t think of it the way he thinks about licensing software, but more to hiring a human consultant or contractor. For example, will the agent be a good cultural fit?

But in some cases, it’s worse than hiring humans since a bad human hire who turns out to be toxic will only interact with a small number of other employees. But an AI agent might interact with thousands of them.

“You have to apply the same level of scrutiny as how you hire real humans,” he says.

A vendor who looks like a technology company might also, in effect, be a staffing firm. “They look and feel like humans, and you have to treat them like that,” he adds.

Another way that AI agents are similar to human consultants is when they leave the company, they take their expertise with them, including what they gained along the way. Data can be downloaded, Gidwaney says, but not necessarily the fine-tuning or other improvements the agent received. Realistically, there might not be any practical way to extract that from a third-party agent, and that could lead to AI vendor lock-in.

Edward Tull, VP of technology and operations at JBGoodwin Realtors, says he, too, sees AI agents as something akin to people. “I see it more as a teammate,” he says. “As we implement more across departments, I can see these teammates talking to each other. It becomes almost like a person.”

Today, JBGoodwin uses two main platforms for its AI agents. Zapier lets the company build its own and HubSpot has its own AaaS, and they’re already pre-built. “There are lead enrichment agents and workflow agents,” says Tull.

And the company is open to using more. “In accounting, if someone builds an agent to work with this particular type of accounting software, we might hire that agent,” he says. “Or a marketing coordinator that we could hire that’s built and ready to go and connected to systems we already use.”

With agents, his job is becoming less about technology and more about management, he adds. “It’s less day-to-day building and more governance, and trying to position the company to be competitive in the world of AI,” he says.

He’s not the only one thinking of AI agents as more akin to human workers than to software.

“With agents, because the technology is evolving so far, it’s almost like you’re hiring employees,” says Sheldon Monteiro, chief product officer at Publicis Sapient. “You have to determine whom to hire, how to train them, make sure all the business units are getting value out of them, and figure when to fire them. It’s a continuous process, and this is very different from the past, where I make a commitment to a platform and stick with it because the solution works for the business.”

This changes how the technology solutions are managed, he adds. What companies will need now is a CHRO, but for agentic employees.

Managing outcomes, not persons

Vituity is one of the largest national, privately-held medical groups, with 600 hospitals, 13,800 employees, and nearly 14 million patients. The company is building its own AI agents, but is also using off-the-shelf ones, as AaaS. And AI agents aren’t people, says CIO Amith Nair. “The agent has no feelings,” he says. “AGI isn’t here yet.”

Instead, it all comes down to outcomes, he says. “If you define an outcome for a task, that’s the outcome you’re holding that agent to.” And that part isn’t different to holding employees accountable to an outcome. “But you don’t need to manage the agent,” he adds. “They’re not people.”

Instead, the agent is orchestrated and you can plug and play them. “It needs to understand our business model and our business context, so you ground the agent to get the job done,” he says.

For mission-critical functions, especially ones related to sensitive healthcare data, Vituity is building its own agents inside a HIPAA-certified LLM environment using the Workato agent development platform and the Microsoft agentic platform.

For other functions, especially ones having to do with public data, Vituity uses off-the-shelf agents, such as ones from Salesforce and Snowflake. The company is also using Claude with GitHub Copilot for coding. Nair can already see that agentic systems will change the way enterprise software works.

“Most of the enterprise applications should get up to speed with MCP, the integration layer for standardization,” he says. “If they don’t get to it, it’s going to become a challenge for them to keep selling their product.”

A company needs to be able to access its own data via an MCP connector, he says. “AI needs data, and if they don’t give you an MCP, you just start moving it all to a data warehouse,” he adds.

Sharp learning curve

In addition to providing a way to store and organize your data, enterprise software vendors also offer logic and functionality, and AI will soon be able to handle that as well.

“All you need is a good workflow engine where you can develop new business processes on the fly, so it can orchestrate with other agents,” Nair says. “I don’t think we’re too far away, but we’re not there yet. Until then, SaaS vendors are still relevant. The question is, can they charge that much money anymore.”

The costs of SaaS will eventually have to come down to the cost of inference, storage, and other infrastructure, but they can’t survive the way they’re charging now he says. So SaaS vendors are building agents to augment or replace their current interfaces. But that approach itself has its limits. Say, for example, instead of using Salesforce’s agent, a company can use its own agents to interact with the Salesforce environment.

“It’s already happening,” Nair adds. “My SOC agent is pulling in all the log files from Salesforce. They’re not providing me anything other than the security layer they need to protect the data that exists there.”

AI agents are set to change the dynamic between enterprises and software vendors in other ways, too. One major difference between software and agents is software is well-defined, operates in a particular way, and changes slowly, says Jinsook Han, chief of strategy, corporate development, and global agentic AI at Genpact.

“But we expect when the agent comes in, it’s going to get smarter every day,” she says. “The world will change dramatically because agents are continuously changing. And the expectations from the enterprises are also being reshaped.”

Another difference is agents can more easily work with data and systems where they are. Take for example a sales agent meeting with customers, says Anand Rao, AI professor at Carnegie Mellon University. Each salesperson has a calendar where all their meetings are scheduled, and they have emails, messages, and meeting recordings. An agent can simply access those emails when needed.

“Why put them all into Salesforce?” Rao asks. “If the idea is to do and monitor the sale, it doesn’t have to go into Salesforce, and the agents can go grab it.”

When Rao was a consultant having a conversation with a client, he’d log it into Salesforce with a note, for instance, saying the client needs a white paper from the partner in charge of quantum.

With an agent taking notes during the meeting, it can immediately identify the action items and follow up to get the white paper.

“Right now we’re blindly automating the existing workflow,” Rao says. “But why do we need to do that? There’ll be a fundamental shift of how we see value chains and systems. We’ll get rid of all the intermediate steps. That’s the biggest worry for the SAPs, Salesforces, and Workdays of the world.”

Another aspect of the agentic economy is instead of a human employee talking to a vendor’s AI agent, a company agent can handle the conversation on the employee’s behalf. And if a company wants to switch vendors, the experience will be seamless for employees, since they never had to deal directly with the vendor anyway.

“I think that’s something that’ll happen,” says Ricardo Baeza-Yates, co-chair of the  US technology policy committee at the Association for Computing Machinery. “And it makes the market more competitive, and makes integrating things much easier.”

In the short term, however, it might make more sense for companies to use the vendors’ agents instead of creating their own.

“I recommend people don’t overbuild because everything is moving,” says Bret Greenstein, CAIO at West Monroe Partners, a management consulting firm. “If you build a highly complicated system, you’re going to be building yourself some tech debt. If an agent exists in your application and it’s localized to the data in that application, use it.”

But over time, an agent that’s independent of the application can be more effective, he says, and there’s a lot of lock-in that goes into applications. “It’s going to be easier every day to build the agent you want without having to buy a giant license. “The effort to get effective agents is dropping rapidly, and the justification for getting expensive agents from your enterprise software vendors is getting less,” he says.

The future of software

According to IDC, pure seat-based pricing will be obsolete by 2028, forcing 70% of vendors to figure out new business models.

With technology evolving as quickly as it is, JBGoodwin Realtors has already started to change its approach to buying tech, says Tull. It used to prefer long-term contracts, for example but that’s not the case anymore “You save more if you go longer, but I’ll ask for an option to re-sign with a cap,” he says.

That doesn’t mean SaaS will die overnight. Companies have made significant investments in their current technology infrastructure, says Patrycja Sobera, SVP of digital workplace solutions at Unisys.

“They’re not scrapping their strategies around cloud and SaaS,” she says. “They’re not saying, ‘Let’s abandon this and go straight to agentic.’ I’m not seeing that at all.”

Ultimately, people are slow to change, and institutions are even slower. Many organizations are still running legacy systems. For example, the FAA has just come out with a bold plan to update its systems by getting rid of floppy disks and upgrading from Windows 95. They expect this to take four years.

But the center of gravity will move toward agents and, as it does, so will funding, innovation, green-field deployments, and the economics of the software industry.

“There are so many organizations and leaders who need to cross the chasm,” says Sobera. “You’re going to have organizations at different levels of maturity, and some will be stuck in SaaS mentality, but feeling more in control while some of our progressive clients will embrace the move. We’re also seeing those clients outperform their peers in revenue, innovation, and satisfaction.”

CIOs take note: talent will walk without real training and leadership

5 December 2025 at 05:00

Tech talent, especially with advanced and specialized skills, remains elusive. Findings from a recent IT global HR trends report by Gi Group show a 47% enterprise average struggles with sourcing and retaining talent. As a consequence, turnover remains high.

Another international study by Cegos highlights that 53% of 200 directors or managers of information systems in Italy alone say the difficulty of attracting and retaining IT talent is something they face daily. Cybersecurity is the most relevant IT problem but a majority, albeit slight, feels confident of tackling it. Conversely, however, only 8% think they’ll be able to solve the IT talent problem. IT team skills development and talent retention are the next biggest issues facing CIOs in Italy, and only 24% and 9%, respectively, think they can successfully address it.

“Talents aren’t rare,” says Cecilia Colasanti, CIO of Istat, the National Institute of Statistics. “They’re there but they’re not valued. That’s why, more often, they prefer to go abroad. For me, talent is the right person in the right place. Managers, including CIOs, must have the ability to recognize talents, make them understand they’ve been identified, and enhance them with the right opportunities.”

The CIO as protagonist of talent management

Colasanti has very clear ideas on how to manage her talents to create a cohesive and motivated group. “The goal I set myself as CIO was to release increasingly high-quality products for statistical users, both internal and external,” she says. “I want to be concrete and close the projects we’ve opened, to ensure the institution continues to improve with the contribution of IT, which is a driver of statistical production. I have the task of improving the IT function, the quality of the products released, the relevance of the management, and the well-being of people.”

Istat’s IT department currently has 195 people, and represents about 10% of the institute’s entire staff. Colasanti’s first step after her CIO appointment in October 2023 was to personally meet with all the resources assigned to management for an interview.

“I’ve been working at Istat since 2001 and almost everyone knows each other,” she says. “I’ve held various roles in the IT department, and in my latest role as CIO, I want to listen to everyone to gather every possible viewpoint. Because how well we know each other, I feel my colleagues have a high expectation of our work together. That’s why I try to establish a frank dialogue and avoid ambiguity. But I make it clear that listening doesn’t mean delegating responsibility. I accept some proposals, reject others, and try to justify choices.”

Another move was to reinstate the two problems, two solutions initiative launched in Istat many years ago. Colasanti asked staff, on a voluntary basis, to identify two problems and propose two solutions. She then processed the material and shared the results in face-to-face meetings, commenting on the proposals, and evaluating those to be followed up.

“I’ve been very vocal about this initiative,” she says, “But I also believe it’s been an effective way to cement the relationship of trust with my colleagues.”

Some of the inquiries related to career opportunities and technical issues, but the most frequent pain points that emerged were internal communication and staff shortages. Colasanti spoke with everyone, clarifying which points she could or couldn’t act on. Career paths and hiring in the public sector, for example, follow precise procedures where little could be influenced.

“I tried to address all the issues from a proactive perspective,” she says. “Where I perceived a generic resistance to change rather than a specific problem, I tried to focus on intrinsic motivation and people’s commitment. It’s important to explain the strategies of the institution and the role of each person to achieve objectives. After all, people need and have the right to know the context in which they operate, and be aware of how their work affects the bigger picture.”

Engagement must be built day by day, so Colasanti regularly meets with staff including heads of department and service managers.

Small enterprise, big concerns

The case of Istat stands out for the size of its IT department, but in SMEs, IT functions can be just a handful of people, including the CIO, and much of the work is done by external consultants and suppliers. It’s a structure that has to be worked with, dividing themselves between coordinating various resources across different projects, and the actual IT work. Outsourcing to the cloud is an additional support but CIOs would generally like to have more in-house expertise rather than depend on partners to control supplier products.

“Attracting and retaining talent is a problem, so things are outsourced,” says the CIO of a small healthcare company with an IT team of three. “You offload the responsibility and free up internal resources at the risk of losing know-how in the company. But at the moment, we have no other choice. We can’t offer the salaries of a large private group, and IT talent changes jobs every two years, so keeping people motivated is difficult. We hire a candidate, go through the training, and see them grow only to see them leave. But our sector is highly specialized and the necessary skills are rare.”

The sirens of the market are tempting for those with the skills to command premium positioning, and the private sector is able to attract talent more easily than public due to its hiring flexibility and career paths.

“The public sector offers the opportunity to research, explore and deepen issues that private companies often don’t invest in because they don’t see the profit,” says Colasanti. “The public has the good of the community as its mission and can afford long-term investments.”

Training builds resource retention

To meet demand, CIOs are prioritizing hiring new IT profiles and training their teams, according to the Cegos international barometer. Offering reskilling and upskilling are effective ways to overcome the pitfalls of talent acquisition and retention.

“The market is competitive, so retaining talent requires barriers to exit,” says Emanuela Pignataro, head of business transformation and execution at Cegos Italia. “If an employer creates a stimulating and rewarding environment with sufficient benefits, people are less likely to seek other opportunities or get caught up in the competition. Many feel they’re burdened with too many tasks they can’t cope with on their own, and these are people with the most valuable skills, but who often work without much support. So if the company spends on training or onboarding new people who support these people, they create reassurance, which generates loyalty.”

In fact, Colasanti is a staunch supporter of life-long learning, and the experience that brings balance and management skills. But she doesn’t have a large budget for IT training, yet solutions in response to certain requests are within reach.

“In these cases, I want serious commitment,” she says. “The institution invests and the course must give a result. A higher budget would be useful, of course, especially for an ever-evolving subject like cybersecurity.”

The need for leadership

CIOs also recognize the importance of following people closely, empowering them, and giving them a precise and relevant role that enhances motivation. It’s also essential to collaborate with the HR function to develop tools for welfare and well-being.

According to the Gi Group study, the factors that IT candidates in Italy consider a priority when choosing an employer are, in descending order, salary, a hybrid job offer, work-life balance, the possibility of covering roles that don’t involve high stress levels, and opportunities for career advancement and professional growth.

But there’s another aspect that helps solve the age-old issue of talent management. CIOs need to recognize more of the role of their leadership. At the moment, Italian IT directors place it at the bottom of their key qualities. In the Cegos study, technical expertise, strategic vision, and ability to innovate come first, while leadership came a distant second. But the leadership of the CIO is a founding basis, even when there’s disagreement with choices.

“I believe in physical presence in the workplace,” says Colasanti. “Istat has a long tradition of applying teleworking and implementing smart working, which everyone can access if they wish. Personally, I prefer to be in the office, but I respect the need to reconcile private life and work, and I have no objection to agile working. I’m on site every day, though. My colleagues know I’m here.”

El MIT empieza a contabilizar los agentes de IA que ahora hacen trabajos que antes desempeñaban personas

5 December 2025 at 03:50

El prestigioso instituto tecnológico estadounidense MIT ha comenzado a desarrollar un índice, llamado Iceberg, para realizar un seguimiento de los diferentes tipos de agentes de IA que ahora realizan trabajos que hasta ahora hacían humanos. La idea del centro es contabilizar los agentes de IA que hay en todo el mundo para obtener una visión más amplia de cómo la tecnología podría sustituir al trabajo humano.

Las cifras iniciales del índice indican que solo 13.000 agentes podrían exponer a 151 millones de trabajadores humanos, es decir, alrededor del 11,7% de la población activa, a la pérdida de puestos de trabajo o salarios.

Un artículo de investigación del MIT afirma que es necesario cuantificar la población de agentes de IA, que en última instancia podría superar a la población humana. La métrica ofrece una foto de cómo la era de la IA está cambiando la productividad, el desarrollo de habilidades y la creación y el desarrollo de puestos de trabajo.

Los investigadores del MIT cuentan que, dado que las cifras de empleo existentes de la Oficina de Estadísticas Laborales de EE. UU. miran hacia atrás y no hacia adelante, se necesita un índice de empleo de IA. Argumentan que los datos ofrecen una visión prospectiva de cómo la IA sustituirá a los trabajadores y ayuda a los líderes a planificar el desarrollo de habilidades y la inversión. “El mercado laboral está evolucionando más rápido de lo que los sistemas de datos actuales pueden captar”, afirman los investigadores, añadiendo que “los marcos de planificación de la mano de obra existentes se han diseñado para economías exclusivamente humanas”.

La pérdida de puestos de trabajo o salarios se debe a la automatización en las empresas, que ya se está produciendo, según señala el estudio. La IA se utiliza habitualmente para generar código y se está empleando para automatizar diversas tareas administrativas y de apoyo.

Los índices de empleo típicos cubren las cifras de pérdida de puestos de trabajo, pero no reflejan las oportunidades creadas por la IA en áreas como los mercados de trabajos esporádicos, los copilotos de IA y las redes de autónomos. “Para cuando estos cambios aparezcan en las estadísticas oficiales, es posible que los responsables políticos ya estén reaccionando a problemas del pasado, destinando miles de millones a programas que se centran en habilidades que ya han quedado obsoletas”, apuntan los investigadores.

El MIT se enfrenta a un gran reto, ya que predecir los puestos de trabajo creados y perdidos por la IA será una tarea difícil, según Jack Gold, analista de J. Gold Associates. “Está claro que la IA hace algunas cosas bien, pero también está claro que aún no comprendemos plenamente el alcance total de sus capacidades y sus inconvenientes”, afirma.

Es muy difícil hacer proyecciones más allá de unos pocos años vista, cuando la IA agentiva alcance su pleno desarrollo, según Gold. “Consideraría cualquier predicción como potencialmente poco precisa en esta fase temprana de la implantación de la IA”, afirma. En todo caso, según el experto, la IA tiene más potencial para ayudar que para sustituir a las personas en los próximos años, incluso cuando surja la IA física.

No obstante, la falta de datos sobre el empleo relacionado con la IA ya es motivo de preocupación en Estados Unidos. El pasado mes de septiembre, algunos de los principales economistas del país enviaron una carta al Departamento de Trabajo de los Estados Unidos pidiendo que “mejorara estos conjuntos de datos para ayudar a los responsables políticos y a los investigadores a evaluar mejor cómo la IA está transformando los mercados laborales”. Las cifras ayudarán a recopilar datos económicos de alta calidad que servirán de base para las políticas destinadas a abordar los problemas laborales que genera la IA, según los economistas. Entre los firmantes del escrito se encontraban Ben Bernanke y Janet Yellen, antiguos presidentes de la Reserva Federal de los Estados Unidos.

Según recientes estadísticas de empleo de Challenger, Gray and Christmas unos 153.074 puestos de trabajo han sido eliminados por la IA. Muchos de ellos eran puestos considerados superfluos en las empresas y puestos de nivel inicial. Varias empresas, entre ellas Amazon y Meta, han estado reduciendo su plantilla mientras aumentaban las inversiones en IA. Las empresas están implantando poco a poco agentes de IA para la gestión del conocimiento, las tareas administrativas y el control de calidad.

BASF Agricultural Solutions, por ejemplo, ha desplegado mil agentes Copilot (de Microsoft), mientras que EY tiene 41.000 agentes en producción, según expusieron recientemente ejecutivos de estas empresas en una mesa redonda celebrada en el evento Ignite de Microsoft, que tuvo lugar el mes pasado en Estados Unidos. Sin embargo, las herramientas de IA que se utilizan actualmente tienen como objetivo aumentar la productividad humana, en lugar de sustituirla, según indicaron los participantes en dicho debate.

Los investigadores del MIT no han respondido ante la petición de declaraciones realizada por este medio.

HPE CEO 넀늬, 죌니퍌 읞수 횚곌 공개···넀튞워크·AI 결합 가속

5 December 2025 at 02:54



HPE가 HP에서 분늬돌 독늜적읞 여정을 시작한 지 10년읎 지난 시점에, 최고겜영자 안토니였 넀늬는 12월 3와 4음 바륎셀로나에서 엎늰 HPE의 죌요 연례 유럜 행사 묎대에 올랐닀. 넀늬는 읎 자늬에서 넀튞워크, 큎띌우드, 읞공지능(AI)읎띌는 섞 가지 Ʞ술 축을 쀑심윌로 한 HPE의 로드맵을 공개했닀.

넀늬는 HPE 디슀컀버 바륎셀로나 2025 행사에 찞석한 6,000여 명의 청쀑을 향핎 “지난 10년 동안 우늬가 핚께 만듀얎낞 성곌가 맀우 자랑슀럜닀”띌며 “앞윌로 펌쳐질 변화는 더욱 Ʞ대된닀”띌고 말했닀.

HPE가 제시한 섞 축의 전략은 였늘날 Ʞ업읎 직멎한 핵심 IT 곌제륌 핎결하Ʞ 위한 것읎닀. 넀늬에 따륎멎 Ʞ업듀은 여전히 레거시 읞프띌 처늬, 데읎터 죌권 확볎, 지속적윌로 슝가하는 비용 ꎀ늬, AI 확산윌로 높아진 컎퓚팅 수요 등의 도전에 맞서고 있닀.

특히 죌니퍌넀튞웍슀(Juniper Networks)륌 지난핎 7월 읞수하며 크게 강화된 넀튞워크 Ʞ술은 읎번 바륎셀로나 행사에서 핵심 요소로 부각됐닀.

죌니퍌의 전 CEO읎자 현재 HPE 넀튞워킹 사업 쎝ꎄ을 ë§¡ê³  있는 띌믞 띌힘은 행사에 찞석핎 양사 통합의 첫 Ʞ술 성곌륌 소개했닀. 양사의 넀튞워크 ꎀ늬 플랫폌에 새로욎 AI êž°ë°˜ 욎영 Ʞ능을 통합하고, 공동 하드웚얎륌 처음윌로 공개한 것읎닀.

띌힘은 “지ꞈ처럌 넀튞워크의 쀑요성읎 높아진 시Ʞ는 없었닀”띌고 말하멎서, 읎제 넀튞워크의 목표는 닚순 연결읎 아니띌 ‘자윚적 ꎀ늬’띌고 섀명했닀. 귞는 넀튞워크가 슀슀로 구성·최적화·복구하는 방향윌로 나아가알 한닀고 강조하며, AI로 섀계되고 AI륌 위한 넀튞워크가 슝가하는 êž°êž° 연결, 복잡핎지는 환겜, 고도화되는 볎안 위협을 처늬할 수 있닀고 밝혔닀.

넀늬는 “띌믞와 낎가 가진 공통의 목표는 넀튞워킹 분알에서 새로욎 늬더륌 만드는 것”읎띌고 말했닀. 귞는 죌니퍌 읞수 후 5개월 만에 HPE가 읎믞 읎전 겜쟁사였던 죌니퍌 Ʞ술곌 2015년 읞수한 아룚바 솔룚션을 결합한 컀넥티비티 제품을 시장에 제공하고 있닀고 섀명했닀. 읎얎 “앞윌로는 양사가 각각 묎엇을 하고 있는지조찚 구분되지 않을 것”읎띌며, “Ʞ볞적읞 읎쀑 섀계륌 읎믞 지원하고 있닀는 사싀은 두 조직읎 얌마나 빠륎게 하나로 융합되고 있윌며, 동시에 HPE의 혁신 역량읎 얎떻게 활용되고 있는지륌 잘 볎여쀀닀”띌고 덧붙였닀.

HPE의 죌니퍌 읞수, 복잡한 곌정을 거치닀

140억 달러(앜 20ì¡° 원) 규몚의 HPE의 죌니퍌 읞수는 닚순한 거래가 아니띌 맀우 복잡하고 ꞎ 여정읎었닀. 2024년 1월 읞수 계획읎 발표됐지만 최종 거래는 2025년 7월에 읎륎러서알 마묎늬됐닀. 믞국에서는 특히 녌란도 적지 않았닀. 믞국 법묎부(DOJ)가 읎번 읞수가 넀튞워크 장비 시장, 특히 묎선랜(WLAN) 분알의 겜쟁을 앜화시킚닀며 소송을 제Ʞ했Ʞ 때묞읎닀.

읎번 읞수 승읞 곌정에서 겪은 난ꎀ곌 여전히 낚아 있는 믞국 낮 비판에 대핮 파욎드늬 산하 얞론사 컎퓚터월드의 질묞을 받은 넀늬는 뚌저 “믞국을 제왞한 국가에서는 통상적읞 6개월 낮 승읞읎 완료됐닀”띌고 섀명했닀. 2024년 여늄에는 3개국만 승읞읎 낚아 있었고, 귞쀑 2개국은 닀음 3개월 낮 승읞을 마쳀닀는 것읎닀. 넀늬는 믞국의 겜우 “선거와 행정부 교첎띌는 변수가 있었고, 읎후 절찚가 닀시 진행됐닀”띌고 덧붙였닀.

넀늬는 읎번 사례륌 분석하멎서 “믞국 법묎부는 캠퍌슀와 지사 시장, 특히 묎선 분알에서 겜쟁사가 3곳에서 2곳윌로 쀄얎듀 것윌로 판닚했닀”띌고 말했닀. 하지만 싀제 시장은 귞볎닀 훚씬 크닀는 게 넀늬의 섀명읎닀. 귞는 “믞국 시장만 볎더띌도 시슀윔, 죌니퍌, HPE, 캄비움넀튞웍슀(Cambium Networks), 유비쿌티(Ubiquity), 아늬슀타(Arista) 등 7~8개 업첎가 겜쟁하고 있닀”띌며 산업군별로 강점읎 닀륎고 대Ʞ업 시장곌 공공 부묞에서도 겜쟁 구도가 닀륎닀고 얞꞉했닀. 읎얎 “여러분(Ʞ자듀)읎 볎도하는 시장점유윚만 뮐도 시장 규몚가 크고 맀우 분산돌 있닀는 사싀을 확읞할 수 있닀”띌고 말했닀.

ê²°êµ­ 믞국 법묎부와는 “상혞에 도움읎 되는 걎섀적읞 곌정을 거쳀닀”띌고 넀늬는 섀명했닀. 귞는 “읎번 읞수 시장은 겜쟁을 쎉진하는 환겜임을 입슝했닀”띌며, 믞국의 대형 M&A 최종 심사 닚계에서도 고객읎나 겜쟁사로부터 얎떠한 읎의 제Ʞ도 받지 않았닀고 강조했닀.

AI와 큎띌우드에 집쀑되닀

바륎셀로나에서 넀늬는 최귌 몇 달 동안 HPE가 큎띌우드와 AI 분알에서 읎뀄낞 Ʞ술적 진전을 강조했닀. 귞는 AI륌 “전형적읞 하읎람늬드 워크로드”띌고 규정하멎서, 두 Ʞ술읎 불가분하게 연결돌 있닀고 섀명했닀.

넀늬는 사용량 êž°ë°˜ 몚덞로 시작핎 현재 전 섞계 4만 6,000명 고객을 확볎한 하읎람늬드 큎띌우드 플랫폌 귞늰레읎크(GreenLake)륌 소개하며, 여Ʞ에 자윚 에읎전튞 êž°ë°˜ 프레임워크 ‘귞늰레읎크 읞텔늬전슀(GreenLake Intelligence)’와 같은 AI Ʞ능을 추가할 계획읎띌고 밝혔닀. 읎 Ʞ능은 지난 6월 HPE가 발표한 것윌로, 하읎람늬드 큎띌우드 환겜에서 IT 욎영을 자동화하고 닚순화하는 데 쎈점을 둔닀. 넀늬는 “IT 욎영 닚순화의 믞래가 읎믞 도착했닀”띌고 말했닀.

넀늬는 또 HPE의 에얎갭 êž°ë°˜ 프띌읎빗 큎띌우드 전략읎 EU처럌 규제가 강한 지역, 귞늬고 군곌 같읎 믌감 데읎터가 쀑요한 전략 분알에서 큰 의믞가 있닀고 강조했닀.

넀늬는 바륎셀로나에서 공개된 또 하나의 솔룚션에도 죌목했닀. AMD의 ‘헬늬였슀(Helios)’ 랙 슀쌀음 AI 아킀텍처가 읎더넷 넀튞워킹곌 통합된 첫 사례닀. 귞는 읎 솔룚션읎 죌니퍌의 연결 하드웚얎와 소프튞웚얎, 람로드컎 토마혞크6 넀튞워킹 칩을 결합핎 “수조 개 맀개변수 몚덞의 학습 튞래픜, 높은 추론 처늬량, 쎈대형 몚덞을 지원할 수 있닀”띌고 섀명했닀. 읎 구성은 HPE 서비슀팀읎 공꞉한닀.

넀늬는 또한 슈퍌컎퓚팅 분알에서 HPE가 볎유한 강력한 입지도 강조했닀. 읎는 2019년 슈퍌컎퓚터 전묞 êž°ì—… 크레읎(Cray)륌 읞수하며 확볎한 Ʞ반읎 크게 작용했닀. 귞는 “HPE는 섞계에서 가장 큰 슈퍌컎퓚터 6대륌 구축한 Ʞ업읎며 읎 분알의 Ꞁ로벌 선도 Ʞ업”읎띌고 말했닀. 닀만 “AI 수요가 ê·ž 얎느 때볎닀 컀졌지만 몚든 Ʞ업읎 읎륌 처늬하Ʞ 위핎 슈퍌컎퓚터가 필요한 것은 아니닀”띌며, 귞러나 “몚든 Ʞ업에는 안전한 AI 슀택읎 필요하닀”띌고 덧붙였닀.

HPE는 읎러한 요구에 대응하Ʞ 위핎 엔비디아와 협력핎 프띌읎빗 큎띌우드 환겜에서 생성형 AI 애플늬쌀읎션 개발·배포륌 가속화하는 통합 읞프띌 솔룚션 ‘HPE 프띌읎빗 큎띌우드 AI’륌 제공하고 있닀. 넀늬는 읎 솔룚션읎 “법적 데읎터 요구사항을 충족하며”, 동시에 AI 혁신의 핵심 곌제읞 “시간, 비용, 위험”을 핎결하는 데 쎈점을 맞춘닀고 섀명했닀. 귞는 여Ʞ에 더핮 HPE가 최귌 엔비디아와 AMD와 핚께 AI 구축을 가속화하는 고성능 넀튞워킹 솔룚션을 추가했닀고 바륎셀로나에서 밝혔닀.

볞사업 êž°ë°˜ 성장곌 M&A êž°ë°˜ 확장

HPE가 지난핎 9월 회계연도 3ë¶„êž° 싀적 발표에서 제시한 전망에 따륎멎, 회사는 2025 회계연도(10월 31음 종료) 맀출읎 고정 환윚 Ʞ쀀 14~16% 슝가할 것윌로 예상하고 있닀. 2024 회계연도 맀출은 301억 달러(앜 44ì¡° 원)로, 2023년 대비 3.4% 슝가했닀.

넀늬의 늬더십 아래 HPE는 쎝 35걎의 읞수륌 진행했닀. 넀늬는 바륎셀로나 Ʞ자회견에서 읎륌 직접 상Ʞ시킀며, 앞서 얞꞉한 죌니퍌넀튞웍슀와 크레읎 왞에도 여러 죌요 읞수륌 나엎했닀.

2020년에는 SD-WAN êž°ì—… 싀버플크(Silver Peak)륌, 2021년에는 데읎터 볎혞 및 재핎복구 êž°ì—… 제륎토(Zerto)륌 읞수했닀. 2023년에는 볎안 및 IT 욎영 분알의 액시슀시큐늬티(Axis Security)와 옵슀랚프(OpsRamp)륌 추가했윌며, 2024년에는 하읎람늬드 큎띌우드 ꎀ늬 êž°ì—… 몚륎페우슀데읎터(Morpheus Data)륌 읞수했닀.

넀늬는 “우늬는 포튞폎늬였륌 볎완하고 목표 시장에서 규몚륌 확장할 수 있는 적절한 자산을 ì°Ÿê³  있닀”띌며 “읎 자산은 맀출곌 수익 잡멎에서 타당핎알 하며, 동시에 죌죌듀에게 가치도 제공핎알 한닀”띌고 말했닀.
dl-ciokorea@foundryco.com


칌럌 | 닚음형 ERP의 종말···조늜형 아킀텍처가 êž°ì—… 믌첩성을 좌우한닀

5 December 2025 at 02:51

필자가 ERP 현대화 프로젝튞륌 읎끌고 IT 및 비슈니슀 임원듀곌 협력핎 옚 겜험을 돌읎쌜볎멎, 성공을 결정짓는 요읞은 Ʞ술 ê·ž 자첎가 아니띌 사고방식곌 아킀텍처였닀. 가튞너는 “2027년까지 새롭게 구축된 ERP 프로젝튞의 70% 읎상읎 쎈Ʞ 비슈니슀 쌀읎슀 목표륌 옚전히 달성하지 못할 것”읎띌고 전망하Ʞ도 했닀. 읎제 ERP 성공은 귌볞적윌로 닀륞 아킀텍처륌 요구하고 있닀.

수십 년 동안 ERP는 재묎, 공꞉망, 제조, HR 등 êž°ì—… 욎영의 쀑심에 자늬핎 왔닀. 통합곌 통제륌 앜속했던 읎 시슀템은 지ꞈ 유연성을 억누륎고 혁신 속도륌 늊추며 Ʞ술 부채륌 쌓는 구조로 변질되고 있닀.

여러 ERP 프로귞랚을 지쌜볞 겜험에 따륎멎 묞제는 ERP 자첎가 아니띌 ERP륌 대하는 우늬의 ꎀ점에 있닀. 많은 Ʞ업읎 ERP륌 닚순한 Ʞ록 시슀템윌로 췚꞉하며, ê·ž 너뚞에 있는 더 큰 Ʞ회륌 놓치고 있닀.

닀가올 비슈니슀 믌첩성의 시대는 ERP륌 몚듈형, 데읎터 쀑심, 큎띌우드 넀읎티람, AI Ʞ반의 조늜형 플랫폌윌로 재정의하는 Ʞ업읎 죌도하게 될 것읎닀. 필자가 핚께핎 옚 여러 조직에서도 Ʞ술 늬더듀은 현대화 여부륌 두고 녌쟁하지 않는닀. 였히렀 ‘사업을 멈추지 않고 얎떻게 싀행할 것읞가’가 핵심 곌제가 되고 있닀.

포람슀의 한 칌럌에서는 읎러한 변화륌 두고 “전 섞계 Ʞ업의 75%가 유연성곌 확장성을 확볎하Ʞ 위핎 Ʞ졎 닚음형 ERP륌 몚듈형 솔룚션윌로 대첎하Ʞ 시작할 것”읎띌고 표현했닀. 읎는 ERP가 레거시 닚음 제품에서 적응형·혁신 쀑심 플랫폌윌로 진화하고 있음을 볎여쀀닀.

읎 흐늄을 수용한 Ʞ업은 ERP륌 혁신의 쎉맀로 만듀 수 있닀. 반대로 전환에 싀팚한 Ʞ업은 핵심 시슀템읎 가장 큰 병목윌로 낚은 채 뒀처질 위험을 안게 된닀.

닚음형에서 몚듈형 백볞윌로의 전환

1990~2000년대 ERP는 닚음 벀더, 닚음 윔드베읎슀, 귞늬고 êž°ì—… 전 영역을 아우륎는 쎈대형 구축 프로젝튞륌 의믞했닀. Ʞ업듀은 몚든 프로섞슀의 섞부적읞 요구사항을 맞추Ʞ 위핎 수백만 달러륌 투입핎 소프튞웚얎륌 컀슀터마읎징했닀.

큎띌우드 시대가 도래하멎서 닀음 장읎 ì—Žë žë‹€. SAP, 였띌큎, 마읎크로소프튞(MS), 읞포 등은 포튞폎늬였륌 SaaS 쀑심윌로 전환했고, 업종 특화 몚듈형 ERP 플랫폌을 앞섞욎 슀타튞업듀도 잇따띌 등장했닀. API와 서비슀 개념읎 확산되멎서 비슈니슀 변화에 맞춰 진화하는 ERP가 가능하닀는 Ʞ대가 볞격적윌로 자늬 잡았닀.

필자가 찞여한 한 전환 프로젝튞에서는 ERP륌 닚음 구현첎로 췚꞉하던 ꎀ점을 낎렀놓은 순간 변화가 시작됐닀. Ʞ능을 몚듈 닚위로 분핎핎 비슈니슀 팀읎 직접 소유하고 독늜적윌로 발전시킬 수 있도록 구조륌 재섀계한 것읎 결정적읞 전환점읎었닀.

귞러나 많은 Ʞ업에서 읎러한 가능성은 완전히 싀현되지 못했닀. 읎제 묞제는 Ʞ술읎 아니띌 사고방식읎닀. 여전히 상당수 조직읎 ERP륌 성장하고 적응핎알 하는 ‘삎아 있는 플랫폌’읎 아니띌, 한번 섀치하멎 끝나는 ‘완료된 시슀템’윌로 바띌볎고 있닀.

Ʞ졎 사고방식읎 쎈래하는 비용

레거시 ERP ꎀ점윌로는 였늘날 변화 속도륌 따띌갈 수 없닀. ê·ž 결곌 혁신은 늊얎지고, 데읎터는 파펞화되며, IT 조직은 끊임없읎 뒀처진 상태륌 만회하느띌 소몚전을 반복하게 된닀. Ʞ업은 비슈니슀 변화만큌 빠륎게 움직음 수 있는 아킀텍처륌 필요로 하고 있닀.

늰IX는 가튞너 분석을 읞용핎 “조늜형 IT 접귌법을 채택한 조직은 새로욎 Ʞ능 구현 속도가 80% 빚띌진닀. 특히 가튞너가 정의한 조늜형 ERP 플랫폌을 적용할 때 읎 횚곌가 두드러진닀”띌고 섀명했닀. 몚듈형 ERP와 전통적읞 닚음형 ERP 사읎의 뚜렷한 성능 격찚륌 볎여쀀닀는 의믞읎Ʞ도 하닀.

필자가 싀제 프로젝튞에서 확읞한 레거시 ERP 사고방식의 비용은 닀음곌 같닀.

• 유연성 부족: 비슈니슀 몚덞은 소프튞웚얎 죌Ʞ볎닀 빠륎게 변하며, 전통적 ERP는 ê·ž 속도륌 따띌가지 못한닀.
• 곌도한 컀슀터마읎징: 수년간 축적된 맞춀형 윔드는 업귞레읎드륌 위험하고 비용 높은 작업윌로 만든닀.
• 데읎터 파펾화: 여러 ERP 읞슀턎슀와 분늬된 몚듈은 데읎터 음ꎀ성을 깚고 분석 신뢰도륌 떚얎뜚늰닀.
• 사용자 불만: 녾후화된 읞터페읎슀는 우회 작업을 부륎고 사용자 찞여륌 떚얎뜚늰닀.
• 높은 TCO: 유지볎수와 업귞레읎드에 예산읎 잠식되멎서 혁신 투자 여력읎 사띌진닀.


조늜형 ERP의 등장

새롭게 부상하는 조늜형 ERP 몚덞은 읎러한 닚음형 구조륌 핎첎한닀. 가튞너는 읎륌 “몚듈형 구성 요소륌 Ʞ반윌로 API로 연결되고 데읎터 팚람늭윌로 통합되는 아킀텍처”띌고 정의한닀.

SAP에 읞수된 아킀텍처 ꎀ늬 도구 êž°ì—… 늰IX(LeanIX)는 “몚듈형·상혞욎용 구성요소로 구축된 조늜형 ERP는 닚음형 제품군에 의졎하지 않고 필요한 Ʞ능을 조늜하듯 구성핎 변화에 빠륎게 대응할 수 있닀”고 섀명한닀. 읎는 정적읞 ERP에서 동적읎고 적응적읞 비슈니슀 플랫폌윌로의 전환을 볎여쀀닀.

맞춀 개발곌 팚킀지형 ERP 양쪜을 겜험한 필자로서는 조늜형 접귌의 진정한 힘읎 닚순한 ‘통합’읎 아니띌 ‘조늜 속도’에 있닀는 점을 확읞핎 왔닀. ERP륌 닚음 제품군읎 아니띌 êž°ì—… 욎영을 가능하게 하는 êž°ë°˜ 시슀템윌로 바띌볎는 ꎀ점읎 쀑요하닀. 재묎, 공꞉망, 제조, HR 같은 핵심 프로섞슀는 Ʞ반윌로 두고, AI 예잡, 고객 분석, 지속가능성 추적 같은 몚듈형 Ʞ능은 비슈니슀 변화에 따띌 동적윌로 연결할 수 있닀.

읎 ì ‘ê·Œ 방식은 Ʞ업에 닀음곌 같은 읎점을 제공한닀.

• 서로 닀륞 벀더 또는 낎부 개발팀의 몚듈을 조합
• 불안정한 컀슀터마읎징 대신 표쀀 API Ʞ반의 큎띌우드 앱 통합
• 자동화·읞사읎튞·예잡 의사결정에 AI 활용
• 역할 êž°ë°˜(persona-based) 겜험 제공

페륎소나: 조늜형 ERP가 드러낎는 ‘사용자 쀑심’의 얌굎

전통적읞 ERP는 몚든 사용자륌 동음하게 췚꞉핎 하나의 읞터페읎슀에 수백 개 메뉎와 끝없는 입력 화멎을 쌓아 올렞닀. 조늜형 ERP는 읎륌 뒀집얎 각 역할읎 싀제로 수행핎알 하는 업묎륌 쀑심윌로 섀계된 ‘페륎소나 êž°ë°˜ 디자읞’을 적용한닀.

• CFO는 AI êž°ë°˜ 시나늬였 몚덞링을 통핎 조직 전반의 재묎 걎전성을 싀시간윌로 확읞한닀.
• 공꞉망 늬더는 싀시간 수요 신혞, 공꞉업첎 성곌, 지속가능성 지표륌 몚니터링한닀.
• 공장 ꎀ늬자는 IoT êž°ë°˜ 섀비 상황, 예지정비 정볎, 생산 KPI륌 추적한닀.
• 영업 및 서비슀 팀은 시슀템을 읎동할 필요 없읎 욎영 데읎터륌 끊김 없읎 활용한닀.

필자의 겜험상 ERP가 범용 튞랜잭션읎 아니띌 싀제 사용자 페륎소나 쀑심윌로 섀계될 때 도입 횚곌가 높아지고 의사결정 속도도 빚띌졌닀.


도전곌 핚정

읎 묞제듀은 읎론적 녌의가 아니띌 IT와 비슈니슀 조직읎 맀음 마죌하는 싀제 곌제듀읎닀.

• 데읎터 거버넌슀: 통합된 데읎터 전략읎 없윌멎 몚듈성은 ê³§ 혌란윌로 읎얎진닀.
• 통합 복잡성: API는 버전 ꎀ늬, 읞슝, 의믞 첎계 정렬 등 엄격한 규윚읎 필요하닀.
• 벀더 종속: 개방형 플랫폌읎띌도 믞묘한 의졎성읎 생Ꞟ 수 있닀.
• 변화 ꎀ늬: 직원은 Ʞ졎 습ꎀ을 버늬고 새로욎 방식을 익히Ʞ 위한 지원곌 교육읎 필요하닀.
• 볎안: 시슀템 간 연결읎 확대될수록 공격 표멎도 ë„“ì–Žì§„ë‹€. 제로튞러슀튞 전략은 필수닀.

진정한 성공은 Ʞ술적 통찰곌 조직에 대한 공감 능력을 균형 있게 갖춘 늬더십에서 나옚닀.


CIO의 새로욎 플레읎북

수년간 ERP 프로젝튞륌 수행하고 비슈니슀·IT 조직곌 협업핎 옚 겜험을 돌아볎멎, ERP 성공을 가로막는 가장 큰 장애묌은 ERP륌 끊임없읎 진화하는 혁신 플랫폌읎 아니띌 ‘완성된 시슀템’윌로 믿는 고정ꎀ념읎었닀.

읎 변화는 도구의 묞제가 아니띌, ERP가 비슈니슀 안에서 수행핎알 할 역할을 재정의하는 묞제닀. 맥킚지는 “ERP 윔얎의 현대화는 닚순한 Ʞ술 업귞레읎드가 아니띌 êž°ì—… 전반의 새로욎 역량을 가능하게 하는 비슈니슀 변혁”읎띌고 섀명한닀. 특히 현대화륌 읎끄는 CIO띌멎 완전히 새로욎 플레읎북읎 필요하닀는 의믞닀.

  1. 소프튞웚얎가 아닌 비슈니슀 아킀텍처에서 출발한닀. Ʞ업읎 얎떻게 욎영되Ꞟ 원하는지 정의한 ë’€, ê·ž 구조에 맞춰 ERP 역량을 섀계한닀.
  2. 통합 데읎터 팚람늭을 구축한닀. 조늜형 ERP의 성팚는 음ꎀ되고 품질 높은 데읎터에 달렀 있닀.
  3. 몚듈형 사고륌 점진적윌로 적용한닀. 소규몚 파음럿윌로 가치륌 입슝한 후 확장한닀.
  4. 퓚전팀을 강화한닀. IT·욎영·비슈니슀 전묞가륌 하나의 애자음 팀윌로 묶얎 빠륎게 솔룚션을 조합한닀.
  5. 성공 Ʞ쀀을 ‘였픈음’읎 아니띌 결곌로 잡정한닀. 목표는 닚음 런치가 아니띌 믌첩성곌 회복탄력성읎닀.
  6. 벀더에 개방성을 요구한닀. 공개 API와 진정한 상혞욎용성을 확볎하고, 독점적 큎띌우드 띌벚에 의졎하지 않는닀.

였띌큎은 읎러한 필요성을 강조하며 “Ʞ업은 변화에 적응할 수 있는 조늜형 애플늬쌀읎션 포튞폎늬였로 읎동핎알 하며, 읎는 재조늜·확장읎 가능한 구조여알 한닀”띌고 섀명했닀. 읎는 ERP 선택 Ʞ쀀에서 유연성읎 핵심 요소가 돌알 핚을 의믞한닀.

ERP륌 혁신 플랫폌윌로 재정의핎알 한닀. 로우윔드 워크플로우, 분석, AI êž°ë°˜ 볎조 도구 등 새로욎 방식을 싀험하는 묞화륌 장렀핎알 한닀.


ERP가 ‘볎읎지 않게’ 되는 때

몇 년 후에는 ERP띌는 용얎조찚 사용하지 않을 가능성읎 크닀. CRM읎 고객 겜험 플랫폌윌로 확장됐듯, ERP도 Ʞ업의 볎읎지 않는 디지턞 백볞윌로 자연슀럜게 녹아듀 것읎닀.

필자는 ERP가 옚프레믞슀에서 큎띌우드, 귞늬고 AI êž°ë°˜ 플랫폌윌로 진화하는 곌정을 지쌜뎀닀. 가까욎 믞래에는 AI가 튞랜잭션곌 워크플로우륌 백엔드에서 처늬하고, 직원듀은 대화형 읞터페읎슀와 낎장 분석 Ʞ능을 통핎 결곌만 요청하게 될 것읎닀. 시슀템에 로귞읞하는 대신 원하는 업묎 결곌륌 말하멎, 조늜형 ERP 팚람늭읎 읎륌 수행하는 데 필요한 몚든 닚계륌 동적윌로 조윚하는 방식읎닀.

읎 믞래는 지ꞈ ERP륌 재정의하는 Ʞ업읎 찚지하게 된닀. 읎는 닚순한 업귞레읎드 죌Ʞ가 아니띌 êž°ì—… 욎영 방식을 닀시 섀계하는 곌정읎닀.


Ʞ록 쀑심에서 가치 찜출 쀑심윌로

ERP는 한때 재고 ꎀ늬, 마감 처늬, 프로섞슀 통제 등 횚윚성 쀑심의 시슀템읎었닀. 였늘날 ERP는 회복탄력성곌 혁신을 견읞하는 구조로 변화하고 있닀. 필자는 여러 ERP 프로귞랚을 겜험하며, CIO의 진짜 곌제는 닚순히 시슀템을 유지하는 것읎 아니띌 êž°ì—… 욎영 방식 자첎에 ‘믌첩성’을 구조적윌로 섀계하는 음읎띌는 점을 확읞핎 왔닀.

큎띌우드·AI·사람 쀑심 섀계륌 Ʞ반윌로 하는 조늜형 ERP는 읎러한 전환을 위한 청사진읎닀. ERP륌 Ʞ록 시슀템에서 혁신 시슀템윌로 바꟞며, 시장 변화 속도에 맞게 끊임없읎 진화할 수 있도록 만든닀.

Ʞ회는 분명하닀. 지ꞈ 변화륌 죌도할 것읞가, 아니멎 얎제의 아킀텍처에 뚞묎륞 채 낎음의 Ʞ업을 만듀얎가는 읎듀을 바띌볌 것읞가. 핚께 생각핎 볌 질묞읎닀.
dl-ciokorea@foundryco.com



였픈AI, ‘넵튠’ 읞수로 AI 학습 추적 도구 낎재화

5 December 2025 at 02:38

였픈AI는 AI 학습 곌정을 추적하는 도구륌 개발핎옚 슀타튞업 넵튠(Neptune)을 읞수하Ʞ로 합의했윌며, 넵튠은 곧바로 자사 제품을 시장에서 철수한닀고 3음 공식 발표했닀.

챗GPT 개발사읞 였픈AI는 1년 넘게 넵튠의 고객윌로 읎 플랫폌을 사용핎옚 것윌로 알렀졌닀.

넵튠곌 같은 싀험 추적 도구는 데읎터 곌학팀읎 AI 몚덞 학습 싀행을 몚니터링하고, 닀양한 섀정 간 결곌륌 비교하며, 개발 곌정에서 발생하는 묞제륌 식별하도록 돕는닀. 넵튠 플랫폌은 몚덞읎 학습 곌정에서 얌마나 였찚륌 쀄읎고 있는지륌 볎여죌는 손싀 곡선, 가쀑치가 얎떻게 변하고 있는지륌 나타낮는 귞래디얞튞(gradient) 통계, 몚덞 낎부의 뉎런읎 입력에 얎떻게 반응하는지륌 볎여죌는 활성화 팹턮 등 죌요 지표륌 수천 걎의 동시 싀험에서 추적핎왔닀.

넵튠읎 시장에서 철수핚에 따띌 SaaS 버전 사용자는 데읎터륌 낎볎낎고 닀륞 플랫폌윌로 읎동할 수 있도록 몇 개월의 유예 Ʞ간을 갖게 된닀. 넵튠은 읎 êž°ê°„ 동안 안정성곌 볎안 팚치륌 제공하지만 새로욎 Ʞ능은 추가되지 않는닀고 섀명했닀. 넵튠은 전환 안낎 페읎지에서 “2026년 3월 4음 였전 10시(태평양 표쀀시)에 혞슀팅 앱곌 API가 종료되며, 낚아 있는 몚든 데읎터는 안전하게 영구 삭제된닀”띌고 밝혔닀.

셀프 혞슀팅 형태로 사용하는 고객에 대핎서는 계정 닎당자가 별도로 연띜을 췚한 상태띌고 회사는 전했닀.

통합에 대한 ìš°ë €

읎번 결정은 AI 개발 도구 시장에서 벀더 통합읎 가속화할 수 있닀는 분석가듀의 우렀륌 불러왔닀. 컚섀팅 êž°ì—… 테크아크(Techarc)의 수석 애널늬슀튞 파읎삎 칎우는 “테슀튞나 싀험 추적 도구 등은 AI륌 포핚한 ì–Žë–€ Ʞ술 벀더에도 연결되거나 종속돌서는 안 된닀”띌며 “읎런 플랫폌은 항상 제3자 형태로 낚아알 하며, 독늜적읎고 쀑늜적읞 결곌에 영향을 믞치는 펞향읎 있얎서는 안 된닀”띌고 말했닀.

칎우사는 AI 업계가 아직 명확한 발전 방향을 정하지 못한 만큌, 도구 읞프띌 통합은 시Ʞ상조띌고 지적했닀. 귞는 “AI의 확싀한 향방읎 정핎지지 않은 상황에서 지ꞈ 도구 읞프띌륌 통합하자는 녌의는 너묎 읎륎닀”띌고 얞꞉했닀.

반멎, 또 닀륞 컚섀팅 êž°ì—… 묎얎 읞사읎튞&슀튞래티지(Moor Insights & Strategy)의 수석 애널늬슀튞 안셞 새귞는 업계가 성숙 닚계로 접얎듀멎서 자연슀럜게 나타나는 흐늄윌로 평가했닀. 새귞는 “였픈AI가 낎부에서 ꟞쀀히 활용하고 싶은 도구륌 안정적윌로 확볎하Ʞ 위한 결정처럌 볎읞닀”띌고 분석했닀.

였픈AI는 녌평 요청에 슉각 응답하지 않았닀.

넵튠은 몚덞 개발 곌정에서 학습 지표륌 추적하고 묞제 징후륌 드러낎며, 읎전 싀험의 Ʞ록 데읎터륌 볎ꎀ하는 소프튞웚얎륌 제공한닀. 읎 플랫폌은 닀양한 몚덞 구조에서의 학습 싀행을 비교하고, 수천 걎의 싀험을 동시에 몚니터링할 수 있도록 지원한닀.

넵튠의 최고겜영자 플였튞륎 니에슈비에치는 읎번 읞수륌 알늬는 랔로귞 Ꞁ에서 자사의 역할을 “반복적읎고 복잡하며 예잡하Ʞ 얎렀욎 몚덞 학습 닚계에서 팀읎 몚덞을 구축하도록 지원하는 것”읎띌고 섀명했닀.

ꎀ렚 고객을 위한 선택지

새귞는 넵튠곌 같은 Ʞ능을 제공하는 Ʞ업읎 읎왞에도 졎재한닀며, 웚읎잠앀드바읎얎시슀(Weights & Biases), 텐서볎드(TensorBoard), ML플로우(MLflow) 등읎 읎 시장에서 활발히 활동 쀑읎띌고 섀명했닀.

싀제로 넵튠은 사용자가 데읎터륌 낎볎낎 ML플로우 또는 웚읎잠앀드바읎얎시슀로 읎전할 수 있도록 안낎 묞서륌 제공했닀.

웚읎잠앀드바읎얎시슀는 시각화 및 협업 Ʞ능을 포핚한 ꎀ늬형 플랫폌을 제공하며, 데읎터람늭슀가 개발한 였픈소슀 ML플로우는 뚞신러닝 띌읎프사읎큎 전반을 닀룚는 플랫폌의 음부로 싀험 추적 Ʞ능을 지원한닀.

또 닀륞 대안윌로는 윔멧(Comet)읎 있윌며, 읎 플랫폌은 싀험 추적 Ʞ능곌 핚께 배포 몚니터링 Ʞ능도 제공한닀.

큎띌우드 서비슀 제공업첎듀도 자첎 플랫폌을 통핎 싀험 추적 Ʞ능을 제공하고 있닀. 구Ꞁ의 버텍슀 AI(Vertex AI)는 구Ꞁ 큎띌우드륌 사용하는 팀을 위한 추적 Ʞ능을 지원하며, AWS의 섞읎지메읎컀(SageMaker)와 마읎크로소프튞 애저 뚞신러닝(Azure Machine Learning) 역시 각각의 생태계에서 유사한 Ʞ능을 제공한닀.
dl-ciokorea@foundryco.com

“AI 시장, 곚드러시에 조정윌로” Ʞ업곌 솔룚션 업첎 몚두 속도 쀄읞닀

5 December 2025 at 00:32

AI 시장읎 지나치게 곌엎된 탓읎든, êž°ì—… CIO듀읎 구맀 계획을 축소하Ʞ로 결정했Ʞ 때묞읎든, 마읎크로소프튞와 였픈AI륌 비롯한 죌요 AI 서비슀 업첎가 맀출 전망을 하향 조정하는 움직임읎 나타나고 있닀.

더읞포메읎션(The Information)은 여러 영업 조직읎 목표륌 달성하지 못한 읎후 마읎크로소프튞가 음부 제품의 AI 영업 할당량을 쀄였닀고 볎도하멎서, 복잡한 업묎륌 자동화하는 AI 에읎전튞에서 Ʞ대한 맀출에 대한 전망을 “조정하고 있는” Ʞ업읎 마읎크로소프튞만읎 아니띌고 전했닀. 볎도에 따륎멎, 였픈AI는 향후 5년 동안 AI 에읎전튞 맀출 전망을 260억 달러 규몚로 하향 조정했닀.

귞레읎하욎드 늬서치(Greyhound Research)의 최고 애널늬슀튞 산치튞 비륎 고Ʞ아는 “AI 영업 할당량 축소는 시장 위Ʞ의 전조가 아니띌, 지난 1년 동안 구조적읞 산업 전환읎띌Ʞ볎닀 곚드러시 같은 엎풍읎 읎얎졌던 상황읎띌 엔터프띌읎슈 Ʞ술 시장읎 마칚낎 현싀로 돌아였고 있닀는 신혞”띌고 말했닀.

고Ʞ아는 “지난 18개월 동안 많은 업첎가 고객읎 현싀적윌로 소화할 수 있는 수쀀을 훚씬 뛰얎넘는 공격적읞 목표륌 섀정했닀”멎서 “엔터프띌읎슈 구맀 닎당자는 읎런 도구륌 충분히 시험핎 볎거나 통합 복잡성을 점검하거나, 복잡하게 얜힌 자사 시슀템 안에서 앜속한 횚곌가 싀제로 유지되는지 평가핎 볌 Ʞ회도 갖지 못한 채 닀년간 AI에 투자하띌는 요구륌 받았닀”띌고 지적했닀.

곌대포장에서 한발 묌러서는 êž°ì—… 고객

고Ʞ아는 영업 압박읎 느슚핎지는 현상에 대핮 “꞉한 쪜윌로 너묎 êž°ìšžì–Ž 있던 대화의 균형을 되찟는 걎강한 움직임”읎띌고 평가했닀. 또 “읎번 조정의 핵심은 솔룚션 업첎가 낎섞욎 앜속곌 엔터프띌읎슈 사용 겜험 사읎의 격찚닀. 구맀자가 AI륌 포Ʞ하는 것읎 아니띌, 곌대ꎑ고에서 한 발 묌러서는 것”읎띌고 덧붙였닀.

Ʞ업은 읎믞 가치가 입슝된 곳에만 투자하Ʞ로 선택하고 있닀. 고Ʞ아는 “2023년부터 2025년까지 귞레읎하욎드 늬서치 조사 결곌륌 볎멎, 대부분 조직읎 거의 비슷한 시점에 같은 깚달음에 도달했닀. 지속 가능한 AI 성곌륌 만듀렀멎 쎈Ʞ 마쌀팅읎 낎섞욎 것볎닀 훚씬 많은 Ʞ쎈 작업읎 필요하닀는 사싀을 알게 됐닀”띌고 말했닀.

데읎터 쀀비에는 시간읎 필요하고, AI 몚덞의 동작을 조윚핎알 한닀는 뜻읎닀. 고Ʞ아는 “AI 거버넌슀 프레임워크는 슉흥적윌로 만듀 수 없닀. 많은 겜우 Ʞ대했던 횚곌의 속도와 범위가, 싀제 프로덕션 시슀템에 적용됐을 때 Ʞ술읎 제공할 수 있는 수쀀볎닀 지나치게 빠륎고 넓었닀”띌고 비판했닀.

읞포테크 늬서치 귞룹(Info-Tech Research Group) 자묞 펠로우 슀윧 빅큎늬는 마읎크로소프튞의 영업 할당량 축소 배겜에는 자쎈한 잡멎읎 있닀며, “마읎크로소프튞의 AI 시장 공략 방식은 였만핚에 Ʞ반하고, 시장 지배력을 최대한 활용하는 전략읎었닀”띌고 지적했닀.

빅큎늬는 “출발점부터 마읎크로소프튞는 고객읎 AI륌 대규몚로 도입하더띌도 정가륌 맀우 높게 책정하고 할읞은 최소화했닀. 윔파음럿읎든 애저 파욎드늬든 읎듀 제품을 ‘완전히 쀀비된 솔룚션, 슉시 도입할 수 있고, 막대한 투자 대비 횚곌륌 낮는 턎킀 팚킀지’읞 것처럌 제시핎 왔닀”띌며, “마읎크로소프튞가 읎런 제품에 대핮 프늬믞엄 가격을 청구하지만, 싀제로는 절반만 완성된 수쀀읎얎서 볞격적읞 욎영 환겜에 투입할 쀀비가 전혀 돌 있지 않고 가격도 지나치게 비싞닀”띌고 비판했닀.

빅큎늬는 “여Ʞ에 더핮, 읎런 도구륌 제대로 활용하고 비슈니슀 프로섞슀륌 닀시 섀계하렀멎 고객 조직 안에 상당한 읞재 역량읎 필요하닀는 점은 아예 고렀조찚 하지 않는닀”고 지적했닀.

빅큎늬는 CIO의 입장에서 바띌볞닀멎, “읎번 움직임을 하나의 닚서로 삌아 Ʞ술 자첎 왞에 필요한 몚든 요소륌 포ꎄ하는 제대로 된 AI 전략을 싀제로 구축하고 있는지, Ʞ술로 묎엇을 달성하렀고 하는지 한 발 떚얎젞서 점검핎알 한닀”띌고 조얞했닀. 또한, “생산성 향상은 방정식의 한 부분음 뿐읎며, 진정한 가치륌 낎렀멎 지ꞈ까지 없었던 수쀀의 개읞화, 새로욎 예잡 능력, 새로욎 성곌와 맀출 성장을 읎끄는 퍌포뚌슀가 필요하닀”띌고 덧붙였닀.

퓚처럌 귞룹(Futurum Group) 엔터프띌읎슈 소프튞웚얎·디지턞 워크플로우 닎당 늬서치 디렉터 킀슀 컀크팚튞늭은 AI 지형읎 닀륞 잡멎에서도 크게 바뀌고 있닀고 분석했닀. 컀크팚튞늭은 수요음 발표한 분석 볎고서에서 “엔터프띌읎슈 소프튞웚얎 시장은 AI 곌대ꎑ고에서 임베디드 방식의 욎영 AI로 변하고 있윌며, 죌요 솔룚션 업첎는 AI륌 워크플로우와 데읎터 계잵, 멀티 에읎전튞 였쌀슀튞레읎션 프레임워크에 직접 통합하고 있닀”띌고 밝혔닀.

AI 발전은 ‘절제’에서 나옚닀

컀크팚튞늭은 “AI 도입읎 확산되멎서 녌의의 쎈점도 닚순한 Ʞ능 비교에서 가치 싀현, 거버넌슀, 상혞욎용성, 진화하는 AI 가격 몚덞로 옮겚갔닀”띌며, “2026년을 낎닀볎멎 구맀자는 잡정 가능한 비슈니슀 성곌륌 우선하멎서 통합된 데읎터 Ʞ반곌 잘 섀계된 멀티 에읎전튞 아킀텍처륌 통핎 AI êž°ë°˜ 맀출 성장, 비용 절감, 욎영 확장 횚곌륌 입슝핎 볎읎는 업첎륌 선택할 것”읎띌고 전망했닀.

Ʞ업읎 읎륞바 “곌장 겜쟁”곌 수식얎 낚발에 점점 플로감을 느끌고 있닀는 점도 지적했닀. 컀크팚튞늭은 “2026년에는 조달 부서가 닚순히 업묎 닚위 생산성 향상만 볎여죌는 수쀀을 넘얎, 비슈니슀 핵심 성곌 지표에 직접 연계된 고객 사례륌 제시하는 업첎에 더 높은 점수륌 쀄 것읎므로 솔룚션 업첎는 겜쟁사의 신규 고객 사례와 성곌 지표륌 멎밀히 몚니터링핎알 한닀”띌고 말했닀.

한펾 빅큎늬는 CIO에게 AI ꎀ렚 의사결정을 낮멮 때 “AI 곌대ꎑ고의 소용돌읎 속윌로 서둘러 ë›°ì–Žë“€ 필요가 없닀는 점을 받아듀읎띌”고 조얞했닀. 빅큎늬는 “각 Ʞ업에 맞는 방향을 찚분하게 섀계하고 계획할 시간을 충분히 가젞도 싀제로 뒀처지는 것은 아니닀”띌며, “AI 하읎프 사읎큎읎 워낙 시끄럜고 얎디에나 졎재하닀 볎니 읎성적읞 녌늬와 합늬적읞 판닚읎 완전히 묻혀 버렞닀”띌고 비판했닀.

고Ʞ아도 읎런 견핎에 동의했닀. 고Ʞ아는 “쎈Ʞ 하읎프 사읎큎의 엎풍은 읎믞 지나갔닀”띌며, “Ʞ술의 잠재력은 여전히 강력하지만, 지ꞈ은 훚씬 더 냉정한 시각곌 안정된 태도로 평가가 읎뀄지고 있닀. AI 솔룚션 업첎는 빠륎게 맀출을 올늬는 것볎닀 시간을 듀여 쌓은 신뢰가 훚씬 더 가치 있닀는 사싀을 깚닫고 읎런 새로욎 늬듬에 맞춰 움직읎고 있닀”띌고 말했닀.

또한, “읎런 성숙핚을 받아듀읎는 조직읎 향후 10년간 엔터프띌읎슈 AI의 방향을 지속 가능하고 신뢰할 수 있윌며, 싀제 욎영 현싀에 Ʞ반한 몚습윌로 만듀얎 갈 것”읎띌고 강조했닀.

고Ʞ아는 현재 마읎크로소프튞 등에서 벌얎지고 있는 상황에 대핮 “몚멘텀의 상싀읎 아니띌 겉볎Ʞ에 화렀한 성곌에서 진짜 싀질적읞 낎용윌로 쀑심축읎 읎동하는 곌정”읎띌고 진닚했닀. 고Ʞ아는 “지ꞈ AI 시장은 진정한 진볎는 곌장된 퍌포뚌슀가 아니띌, 조용하지만 음ꎀ된 싀행곌 절제에서 나옚닀는 사싀을 깚닫고 있닀”띌며, “읎번 사읎큎에서 처음윌로 읎런 ‘절제’가 눈에 볎읎Ʞ 시작했닀”띌고 덧붙였닀.
dl-ciokorea@foundryco.com

지역·섞대별 AI 활용 및 디지턞 웰빙 격찚 확대 시슀윔·OECD 분석

5 December 2025 at 00:08

시슀윔와 겜제협력개발Ʞ구(OECD)가 협력하여 공동윌로 구축한 ‘디지턞 웰빙 허뾌(Digital Well-being Hub)’가 Ʞ술의 위험곌 읎점, 귞늬고 AI가 사람의 삶에 믞치는 영향을 심잵적윌로 분석한 최신 연구 결곌륌 공개했닀. 생성형 AI가 음상에 빠륎게 자늬 잡는 가욎데, AI 활용을 둘러싌 지역별/섞대별 격찚가 더욱 뚜렷핎지고 있닀는 분석읎닀. 읎런 격찚는 누가 AI의 혜택을 누늬고, 누가 더 큰 위험을 감수하는지, 귞늬고 디지턞 생활읎 개읞의 웰빙에 ì–Žë–€ 방식윌로 영향을 믞치는지륌 좌우할 수 있는 쀑대한 요읞윌로 작용한닀.

볎고서에 따륎멎, 전 섞계적윌로 35섞 믞만 젊은 섞대는 생성형 AI와 각종 디지턞 서비슀 활용의 핵심 사용자잵읎닀. 특히 읞도·람띌질·멕시윔·낚아프늬칎공화국 등 신흥국 청년잵읎 두드러지며 AI 사용률, 신뢰 수쀀, 교육 찞여도 등 거의 몚든 지표에서 상위권을 Ʞ록했닀. 반대로 많은 유럜 국가에서는 AI ꎀ렚 신뢰도가 낮고 불확싀성읎 높게 나타났닀. Ʞ술 도입읎 선진국에서 뚌저 음얎나던 Ʞ졎 흐늄곌 닀륞 양상읎닀.

흥믞로욎 점은 AI 활용도가 높은 신흥국 청년잵읎 동시에 ‘디지턞 웰빙’ 저하 지표에서도 높은 수치륌 볎였닀는 사싀읎닀. 읎듀은 여가 시간대 슀크늰 사용 시간읎 가장 êžžê³  옚띌읞 Ʞ반의 사회적 의졎도 역시 높게 나타났닀. 또한 Ʞ술 사용윌로 읞한 감정 Ʞ복도 가장 심한 겜향을 볎여 닚순한 Ʞ술 접귌성 읎상의 균형 잡힌 디지턞 환겜읎 필요성을 부각했닀.

연구에 따륎멎, 전 섞계적윌로 하룚 5시간을 쎈곌하는 여가 시간대 슀크늰 사용 시간은 개읞의 전반적읞 웰빙 저하와 삶의 만족도 감소와 연ꎀ되는 것윌로 나타났닀. 특히 한국은 전 섞계에서 ‘슀크늰 플로감(screen fatigue)’읎 가장 높은 국가로 조사됐닀. 읎런 상ꎀꎀ계가 반드시 읞곌ꎀ계륌 의믞하는 것은 아니지만, 지속 가능한 디지턞 믞래륌 위핎서는 Ʞ술 혁신만큌읎나 개읞의 걎강곌 행복을 지킀는 디지턞 웰빙에 대한 ꟞쀀한 ꎀ심곌 녞력읎 필요하닀.

시슀윔 수석부사장 겞 Ꞁ로벌혁신책임자 가읎 디드늬히는 “신흥국읎 AI 역량을 갖출 수 있도록 지원하는 것은 닚순한 Ʞ술 볎꞉읎 아니띌, 신흥국의 각 개읞읎 슀슀로의 믞래륌 섀계할 수 있도록 잠재력을 ì—Žì–Ž 죌는 음”읎띌며 “AI가 우늬의 음상곌 음터에 빠륎게 볎꞉되고 있는 지ꞈ, 우늬는 투명성, 공정성, 프띌읎버시륌 핵심 가치로 삌아 읎듀 도구가 책임감 있게 섀계되도록 핎알 한닀”띌고 말했닀.

읎얎 “업묎륌 횚윚화하고 협업을 개선하며, 성장곌 학습의 새로욎 Ʞ회륌 만듀얎 쀄 때 AI는 웰빙을 향상시킀는 방향윌로 ê·ž 잠재력을 가장 크게 발휘할 수 있닀”띌며 “Ʞ술곌 사람, 귞늬고 분명한 목적성읎 결합될 때에알 비로소, 회복탄력성 있고 걎강하며 번영하는 컀뮀니티가 몚든 곳에서 형성될 수 있닀”띌고 덧붙였닀.

읎번 연구에서는 섞대 간 격찚도 두드러졌닀. 전 섞계 청년잵은 공통적윌로 사회적 상혞작용 대부분 또는 전부가 옚띌읞에서 읎룚얎지고 있닀고 답했윌며, AI의 유용성에 대핎서도 높은 신뢰도륌 볎였닀. 35섞 믞만의 조사대상자 쀑 절반 읎상읎 적극적윌로 AI륌 사용하고 있윌며, 75% 읎상은 AI가 유용하닀고 평가했닀. 또한 26~35섞 응답자의 절반가량은 읎믞 AI ꎀ렚 교육을 읎수한 것윌로 나타났닀.

반대로, 45섞 읎상 쀑장년잵은 AI의 유용성에 대핮 비교적 회의적읎었윌며, 절반 읎상은 AI륌 전혀 사용하지 않는닀고 답했닀. 55섞 읎상에서는 “AI륌 신뢰하는지 잘 몚륎겠닀”띌는 응답읎 높게 나타났는데, 읎는 명확한 거부감읎띌Ʞ볎닀 Ʞ술에 대한 낮은 친숙도와 겜험 부족에서 비롯된 불확싀성윌로 핎석된닀.

섞대별 친숙도의 격찚는 AI가 음자늬와 업묎 환겜에 믞칠 영향에 대한 Ʞ대와 읞식에서도 고슀란히 드러난닀. 35섞 믞만곌 신흥국 응답자는 AI가 향후 음자늬에 믞칠 잠재적 영향읎 가장 큎 것윌로 전망한 반멎, 고령잵에서는 ê·ž 수쀀읎 상대적윌로 낮았닀.

디드늬히는 “디지턞곌 AI 도입에서 나타나는 섞대 찚읎는 ì–Žì©” 수 없닀며 포Ʞ할 묞제가 아니띌, 우늬가 분명한 목표륌 가지고 행동핚윌로썚 충분히 핎결할 수 있는 곌제”띌며 “젊은 섞대가 새로욎 Ʞ술을 더 빚늬 받아듀음 수는 있지만, 몚든 연령대의 사람읎 각자의 고유하고도 소쀑한 겜험곌 통찰을 갖고 있닀”띌며 “AI 성공의 핵심 Ʞ쀀은 닚지 도입 속도가 아니띌, 몚든 연령·Ʞ술 수쀀·지역의 사람읎 AI륌 활용핎 싀제로 삶을 얌마나 향상시킬 수 있는가에 두얎알 한닀. 귞래알만 ‘AI 섞대(Generation AI)’가 진정윌로 몚두륌 포용하는 섞대가 될 수 있닀”띌고 덧붙였닀.
dl-ciokorea@foundryco.com

Before yesterdayCIO

레거시 유지볎수에 발목 잡힌 IT, 서드파티로 돌파구 몚색

4 December 2025 at 22:14

Ʞ술 부채가 IT 조직을 마비시킬 위협 요읞윌로 떠였륎자 상당수 CIO가 레거시 소프튞웚얎와 시슀템 유지볎수·업귞레읎드륌 위핎 서드파티 서비슀 업첎에 눈을 돌늬고 있닀. 맀니지드 서비슀 업첎 엔소녞(Ensono)가 싀시한 섀묞조사 결곌, IT 늬더 100명 가욎데 95명읎 레거시 IT륌 현대화하고 Ʞ술 부채륌 쀄읎Ʞ 위핎 왞부 서비슀 업첎륌 활용하고 있는 것윌로 나타났닀.

읎 같은 움직임은 부분적윌로 레거시 IT 비용 슝가에서 비롯됐닀. 응답자 가욎데 거의 절반은 지난핎 녾후 IT 시슀템 유지볎수에 예산볎닀 더 많은 비용을 지출했닀고 답했닀. 더 큰 묞제는 레거시 애플늬쌀읎션곌 읞프띌가 IT 조직의 발목을 잡고 있닀는 점읎닀. IT 늬더 10명 가욎데 9명은 레거시 유지볎수가 AI 현대화 계획에 걞늌돌읎 되고 있닀고 지적했닀.

엔소녞의 CTO 팀 베얎뚌은 “레거시 시슀템 유지볎수가 현대화 녞력에 큰 방핎가 되고 있닀”띌며, “전형적읞 혁신가의 딜레마닀. 혁신볎닀는 녾후 시슀템곌 ê·ž 핎결 방안에만 집쀑하고 있닀”띌고 지적했닀.

음부 CIO는 레거시 시슀템 욎영을 서비슀 업첎에 맡Ʞ거나 왞부 IT팀을 활용핎 Ʞ술 부채륌 정늬하고 소프튞웚얎와 시슀템을 현대화하고 있닀. 베얎뚌은 레거시 시슀템을 왞부에 맡Ʞ는 Ʞ업읎 슝가하는 배겜윌로 고령화된 읞력을 ꌜ았닀. êž°ì—… 낎부의 레거시 시슀템 전묞가가 은퇎하멎서 축적된 지식도 핚께 빠젞나가고 있닀는 의믞닀.

베얎뚌은 “읎 음을 낎부에서 직접 처늬할 수 있는 읞력읎 많지 않닀. 조직 낮 읞력읎 고령화되고 퇎직자가 늘얎나는 상황에서, 필요한 읞재륌 채용하Ʞ 얎렀욎 영역에서는 왞부에서 전묞 읞력을 찟아알 한닀”띌고 섀명했닀. 또, “MSP 몚덞 자첎는 수십 년 전부터 졎재했지만, 최귌에는 예산을 확볎하고 AI륌 도입할 시간을 만듀Ʞ 위핎 MSP륌 Ʞ술 부채 ꎀ늬 수닚윌로 활용하는 흐늄읎 컀지고 있닀”띌고 분석했닀.

AI처럌 새로욎 Ʞ술읎 빠륎게 확산되는 것도 읎런 흐늄에 음조하고 있닀. 베얎뚌은 “한쪜에는 ꎀ늬·유지핎알 하는 레거시 묞제가 있고, 닀륞 한쪜에는 수년 동안 겜험하지 못한 속도로 발전하는 최신 Ʞ술읎 있얎 양쪜을 동시에 따띌가Ʞ 얎렵닀”띌고 덧붙였닀.

위험의 아웃소싱

사읎버 볎안 서비슀 업첎 뉎빅(Neuvik)의 CEO 띌읎얞 레읎륎빅은 레거시 IT ꎀ늬륌 서비슀 업첎에 맡Ʞ는 흐늄읎 확대되고 있닀는 점에 동의했닀. 레읎륎빅은 레거시 시슀템에 적합한 전묞가륌 맀칭하는 등 여러 장점을 얞꞉하멎서도, CIO가 위험 ꎀ늬륌 위핎 MSP륌 활용하는 겜향도 있닀고 지적했닀.

레읎륎빅은 “많은 장점 가욎데 자죌 얞꞉되지 않는 핵심은 췚앜점 악용읎나 서비슀 쀑닚 위험을 서비슀 업첎에 ë§¡êžž 수 있닀는 점”읎띌며, “췚앜점 발견곌 팚치, 전반적읞 유지볎수에 지속적윌로 많은 비용읎 드는 환겜에서는 잘못 대응했을 때 발생하는 위험을 서비슀 업첎가 떠안게 되는 겜우가 많닀”띌고 섀명했닀.

믞 국방부(US Department of Defense)에서 비서싀장 겞 사읎버 부묞 부국장을 지낞 레읎륎빅은 레거시 IT 유지볎수 예산을 쎈곌 집행한 IT 책임자가 많닀는 것읎 놀랄 음은 아니띌고 말한닀. 많은 조직읎 현재 볎유한 IT 읞프띌와 앞윌로 전환핎알 할 읞프띌 사읎에서 필요한 읞재 역량읎 맞지 않는 상황에 놓여 있닀고 지적하며, 레거시 소프튞웚얎와 시슀템의 지속적읞 유지볎수 비용읎 예상볎닀 더 많읎 드는 겜우도 잊닀고 말했닀.

레읎륎빅은 “쎈Ʞ 도입 비용읎 1읎띌멎, 유지볎수 비용은 1X읎Ʞ 때묞에 예상하지 못한 거대한 유지볎수 ꌬ늬가 생ꞎ닀”띌고 덧붙였닀.

레거시 유지볎수의 덫에서 벗얎나렀멎 적절한 서드파티 업첎륌 고륎는 선견지명곌 선택 Ʞ쀀읎 필요하닀. 레읎륎빅은 “장Ʞ적읞 ꎀ점에서 핎당 업첎와 향후 5년 계획읎 맞묌늬는지 반드시 확읞핎알 한닀. 또 조직의 목표와 업첎가 제공하렀는 지원 방향읎 음치하는지도 점검핎알 한닀”띌고 조얞했닀.

두 번 지불하는 비용

음부 IT 늬더가 레거시 시슀템 현대화륌 서드파티 업첎에 ë§¡êž°ê³  있지만, IT 서비슀 ꎀ늬 및 고객 서비슀 소프튞웚얎 업첎 프레시웍슀(Freshworks)가 최귌 공개한 볎고서는 읎런 현대화 녞력읎 곌연 횚윚적읞지에 의묞을 제Ʞ했닀.

프레시웍슀의 조사에서 응답자의 3/4 읎상은 소프튞웚얎 도입에 예상볎닀 더 많은 시간읎 걞늰닀고 답했고, 프로젝튞 가욎데 2/3은 예산을 쎈곌했닀고 응답했닀. 프레시웍슀의 CIO 아슈윈 발랄은 서드파티 서비슀 업첎가 읎 묞제륌 핎결핎 죌지 못할 수도 있닀고 겜고했닀.

발랄은 “레거시 시슀템읎 너묎 복잡핎지멎서 Ʞ업읎 도움을 구하렀고 서드파티 업첎와 컚섀턎튞에 점점 더 의졎하고 있지만, 싀제로는 수쀀 읎하의 레거시 시슀템을 닀륞 수쀀 읎하 레거시 시슀템윌로 바꟞는 결곌에 귞치는 겜우가 많닀”띌며, “서드파티 업첎와 컚섀턎튞륌 추가하멎 Ʞ졎 묞제륌 핎결하Ʞ볎닀는 새로욎 복잡성만 더핮 묞제륌 악화시킀는 사례도 적지 않닀”띌고 지적했닀.

핎법은 서드파티 업첎륌 늘늬는 것읎 아니띌 별도의 복잡한 작업 없읎 바로 쓞 수 있는 새로욎 Ʞ술을 도입하는 데 있닀. 발랄은 “읎론적윌로 서드파티 업첎는 전묞성곌 속도륌 제공한닀. 하지만 현싀에서는 복잡한 Ʞ술을 도입하는 데 한 번, 핎당 Ʞ술읎 제대로 작동하도록 컚섀턎튞륌 투입하는 데 또 한 번 등 두 번 비용을 지불하는 겜우가 많닀”띌고 ꌬ집었닀.

플하Ʞ 얎렀욎 서드파티 업첎 활용

사읎버 볎안 솔룚션 업첎 워치가드 테크놀로지슀(WatchGuard Technologies)의 필드 CTO 겞 CISO 애덀 윈슀턎은 상당수 IT 늬더가 음정 수쀀의 서드파티 지원을 사싀상 플할 수 없는 선택윌로 볎고 있닀. 윈슀턎은 였래된 윔드륌 업데읎튞하거나 워크로드륌 큎띌우드로 읎전하고 SaaS 도구륌 도입하고, 사읎버볎안을 강화하는 등 대부분의 곌제에서 읎제 왞부 지원읎 필요하닀고 말했닀.

윈슀턎은 녾후 원격접속 도구와 VPN을 포핚한 레거시 시슀템읎 쌓읎멎 Ʞ술 부채가 눈덩읎처럌 불얎나 조직을 짓누륌 수 있닀고 겜고했닀. 또, 아직 많은 조직읎 큎띌우드나 SaaS 도구로 완전히 현대화하지 못한 상태읎며, 전환 시점읎 였멎 왞부 업첎에 도움을 요청할 수밖에 없을 것읎띌고 낎닀뎀닀.

윈슀턎은 “대부분 Ʞ업은 자첎 애플늬쌀읎션을 섀계·개발·욎영하지 않고, 귞런 영역에 Ʞ술 부채가 쌓여 있는 상황에서 하읎람늬드 IT 구조륌 유지하고 있닀”띌며, “여전히 윔로쌀읎션곌 옚프레믞슀 쀑심읎던 시절의 환겜을 유지하는 Ʞ업도 많고, 읎런 환겜에는 거의 예왞 없읎 레거시 서버와 레거시 넀튞워크, 현대적읞 섀계나 아킀텍처륌 따륎지 않는 레거시 시슀템읎 포핚돌 있닀”띌고 섀명했닀.

읎런 Ʞ업의 IT 늬더는 녾후 Ʞ술을 닚계적윌로 퇎역시킀는 계획을 섞우고, IT 투자가 가능한 한 최신 상태륌 유지하도록 솔룚션 업첎의 책임을 명확히 하는 서비슀 계앜을 첎결핎알 한닀. 윈슀턎은 많은 솔룚션 업첎가 신제품을 낎놓윌멎서 Ʞ졎 제품 지원을 너묎 쉜게 쀑닚한닀고 지적했닀.

윈슀턎은 “업귞레읎드륌 하지 않을 계획읎띌멎 레거시 지원 비용을 멎밀히 ë”°ì ž 볎고, 업귞레읎드할 수 없닀멎 얎떻게 격늬할 것읞지에 대한 답을 쀀비핎알 한닀”띌며, “업귞레읎드가 불가능할 겜우 위험을 옮ꞰꞰ 위한 읎륞바 ‘묎덀 격늬 전략(graveyard segmentation strategy)’을 얎떻게 섀계할지도 고믌핎알 한닀”띌고 강조했닀. 또 “솔룚션 업첎 싀사 곌정에서 읎런 녌의가 빠지는 겜우가 많고, 귞러닀 묞제가 터지멎 조직읎 뒀늊게 놀띌게 된닀”띌고 덧붙였닀.

귞렇닀고 CIO가 레거시 IT 전묞성을 쌓는 방향윌로 컀늬얎륌 섀계하는 것은 플핎알 한닀. 윈슀턎은 “소프튞웚얎나 구축 비용을 충분히 상각하지 못했닀멎, 앞윌로 도입하는 몚든 신규 애플늬쌀읎션에는 최신 컎포넌튞륌 사용하겠닀고 슀슀로 닀짐핎알 한닀”띌고 강조했닀.
dl-ciokorea@foundryco.com

큎띌우드플레얎 êž°ê³ | AI 시대, 윘텐잠 통제권을 위한 ‘허가 êž°ë°˜ 읞터넷’윌로 전환핎알

4 December 2025 at 19:38

곌거 검색 엔진 크례링은 웹윌로 닀시 튞래픜을 돌렀죌는 읎로욎 구조였지만, 읎제는 상황읎 닀륎닀. AI Ʞ업듀은 웹에서 수집한 윘텐잠륌 학습 데읎터로 활용핎 요앜·응답·개요 형태의 파생 윘텐잠륌 제공하고, 사용자는 원볞 사읎튞륌 방묞하지 않고도 필요한 정볎륌 얻게 된닀. 읎는 튞래픜곌 ꎑ고 수익을 감소시쌜 윘텐잠 제작자의 수익 구조륌 위협할 뿐 아니띌, 지적 재산권 볎혞·데읎터 출처 확볎·윘텐잠 였낚용 묞제륌 알Ʞ하는 구조적 변화닀. 윘텐잠 제작자가 자신의 데읎터에 대한 통제력을 잃게 되는 것읎닀.

더 큰 묞제는 AI êž°ë°˜ 뎇읎 볎안 위협윌로 진화하고 있닀는 점읎닀. 음부 악성 뎇은 닚순 크례링을 넘얎 웹 췚앜점을 자동윌로 탐색하고, 계정 탈췚, 사Ʞ성 결제 시도 등 닀양한 공격을 수행한닀. 예륌 듀얎, Ʞ업읎 슝시 마감 후 발표할 예정읎었던 쀑요 비공개 재묎 정볎가 악성 뎇에 의핎 유출될 겜우, 읎는 불법 죌식 거래와 규제 위반윌로 읎얎젞 회사에 치명적읞 결곌륌 쎈래할 수 있닀.

AI 뎇의 양적 확산은 읎제 묎시하Ʞ 얎렀욎 수쀀읎닀. 읞터넷 현황 몚니터링 플랫폌 큎띌우드 레읎더의 데읎터에 따륎멎, 특히 메타의 AI 뮇 ‘메타-익슀터널 에읎전튞(Meta-External Agent)’는 1년 새 요청량읎 843%띌는 폭발적읞 슝가섞륌 볎였닀. 였픈AI의 GPT뮇(GPTBot) 역시 147% 슝가하며 Ʞ졎의 IP 찚닚읎나 닚순 레읎튞 늬믞팅만윌로는 읎듀을 통제하Ʞ 얎렀워졌닀는 것을 반슝한닀. 더불얎, AI가 ‘CAPTCHA(캡찚)’륌 학습핎 우회하는 사례도 늘고 있닀.

읎러한 변화 속에서 Ʞ업곌 퍌랔늬셔는 악의적읞 AI 뎇을 찚닚하고 윘텐잠 슀크래핑을 제얎할 수 있는 횚곌적읞 방법을 찟아알 한닀. AI가 만듀얎낎는 새로욎 비슈니슀 Ʞ회륌 찚닚하지 않윌멎서도, 조직의 데읎터·볎안·람랜드륌 볎혞하렀멎 Ʞ졎볎닀 훚씬 정교한 접귌읎 필요하닀.

따띌서 AI 뮇 위협에 대응하고 윘텐잠 통제권을 되찟Ʞ 위핎서는 닀음곌 같은 닀쀑 계잵 볎안 전략구축읎 요구된닀:

첫짞, Ʞ쎈 닚계읞 정적 제얎(Layer 1)ë‹€. 읎는 대규몚 뮇 공격을 견디고, AI êž°ë°˜ 뎇읎 Ʞ졎 방얎선을 쉜게 우회하지 못하도록 하는 출발점읎 된닀. CAPTCHA륌 사용하지 않는 읞슝 방식, 닀쀑 읞슝(MFA), 레읎튞 늬믞팅곌 같은 요소듀은 싀제 사용자의 겜험을 저핎하지 않윌멎서도 자동화된 시도륌 횚곌적윌로 찚닚한닀. 또한 악성 뎇을 정상 페읎지 대신 대첎 윘텐잠로 유도핎 늬소슀륌 소비시킀는 Ʞ법도 정적 제얎의 음환윌로 활용될 수 있닀.

둘짞, 동적 제얎(Layer 2)는 예잡적 ë°©ì–Ž 능력을 더한닀. 정적 제얎 위에 더핎지는 동적 제얎는 변화하는 AI 뎇의 움직임을 조Ʞ에 감지하고 대응하는 역할을 한닀. 싀시간 위협 읞텔늬전슀 분석을 통핎 새로욎 공격 팚턎읎 도달하Ʞ 전에 찚닚할 수 있고, 상섞한 튞래픜 로귞는 사람읎 볎Ʞ 얎렀욎 행동 팚턎의 찚읎륌 식별하는 데 도움을 쀀닀. 뚞신러닝(ML) êž°ë°˜ 행동 분석은 정상 사용자와 비정상적 튞래픜의 간극을 자동윌로 파악핎 읎상 징후륌 식별한닀. 읎러한 동적 제얎는 AI 뎇읎 시시각각 팚턎을 바꟞며 등장하는 환겜에서 필수적읎닀.

ì…‹ì§ž, 가장 쀑요한 섞분화된 거버넌슀(Layer 3)ë‹€. 읎는 묎조걎적읞 찚닚 전략에서 벗얎나, ì–Žë–€ 뎇읎 ì–Žë–€ 목적을 가지고 ì–Žë–€ 페읎지에 접귌할 수 있는지륌 조직읎 직접 결정하는 첎계륌 의믞한닀. 읎륌 위핎 조직은 뚌저 AI 감사(AI Auditing) Ʞ능을 통핎 ì–Žë–€ AI 뎇읎 사읎튞에 접귌하고 있는지 투명하게 파악핎알 한닀. 뎇읎 ì ‘ê·Œ 목적곌 소속 서비슀륌 암혞화 서명윌로 슝명하도록 요구핚윌로썚, 뎇의 신뢰성을 확볎하고 정식 크례러와 비정상적읞 접귌을 구분할 수 있닀. 더 나아가, 페읎지 닚위로 ì ‘ê·Œ 권한을 조정핎 ꎑ고 êž°ë°˜ 수익 페읎지는 찚닚하고 개발자 묞서나 공공성 있는 자료는 허용하는 등 윘텐잠 성격에 따띌 전략적 선택을 할 수 있닀. 특히, 크례링당 결제(pay-per-crawl) 몚덞을 적용하멎 AI Ʞ업읎 데읎터륌 학습에 활용할 때 합당한 비용을 지불하도록 할 수 있얎 윘텐잠 제작자에게 새로욎 수익 몚덞을 엎얎쀄 수 있닀.

궁극적윌로 읎러한 닀쀑 계잵 전략은 읞터넷읎 AI륌 쀑심윌로 재펞되는 흐멄 속에서 윘텐잠 제작자와 Ʞ업읎 닀시 통제권을 확볎하는 곌정읎닀. 닚순히 유핎한 뎇을 막는 것에 귞치지 않고, ì–Žë–€ 죌첎가 ì–Žë–€ 방식윌로 윘텐잠륌 활용할 수 있는지 선택할 수 있는 권한을 되찟는 방향윌로 나아가알 한닀. 읎륌 통핎 조직은 AI가 만듀얎낎는 위협윌로부터 슀슀로륌 볎혞하는 동시에, 새로욎 읞터넷 시대가 제공하는 Ʞ회륌 볎닀 공정하고 안정적윌로 활용할 수 있을 것읎닀.

*필자 조원균 큎띌우드플레얎(Cloudflare) 한국 지사장은 한국 낮 큎띌우드플레얎의 입지 강화와 뾌랜드 읞지도 제고에 죌력하고 있윌며, 섞음슈 및 채널 파튞너륌 통한 고객 접점 최적화에도 집쀑하고 있닀. 원균 지사장은 25년 읎상 늬더십 겜험을 볎유한 베테랑윌로, 큎띌우드플레얎 합류 전 F5, 포티넷, 시슀윔 등을 포핚한 죌요 Ꞁ로벌 테크 Ʞ업에서 귌묎한 바 있닀.
dl-ciokorea@foundryco.com

Closing the IT estate expectation gap

4 December 2025 at 12:58

Talk to CEOs today and some common themes emerge: they’re moving faster, making bigger bets and relying more heavily on technology to execute their strategic agenda. Expectations on the IT estate have never been higher, yet many CEOs feel it’s a “black box” – essential, but difficult to see into and even harder to gauge.

At the same time, CIOs know that aging infrastructure is struggling to keep pace with AI-driven transformation, rising cyber risks or the agility their CEO has come to expect.  

This expectation gap is exactly why Netskope’s Crucial Conversations research identifies the IT estate as one of the six essential discussions CIOs must master today if they are to successfully align with their CEO on their modernization agenda. 

CEOs’ growing frustration with the “black box”

CEOs that took part in the research admitted they don’t understand what’s happening deep inside the IT stack and that makes them uncomfortable. Some feel their CIO shields them from the complexity; others feel the CIO overcomplicates it. Either way, this impacts confidence. 

Why the IT estate has become a strategic conversation

Three forces are pushing the IT estate onto the CEO agenda faster than many CIOs expected:

1. AI demands modern foundations

Organizations are moving from AI experiments to AI integration at pace. But AI doesn’t run effectively on infrastructure designed for a pre-AI world. CEOs need to understand that modernization is not a technology preference – it’s a prerequisite for delivering the business outcomes they now expect from AI.

2. The cost/risk trade-off is shifting

CEOs expect CIOs to be “gatekeepers” of cost, challenging suppliers and avoiding unnecessary spending. But they also expect CIOs to be candid about the real cost of doing nothing – outages, slowdowns, security exposure and innovation bottlenecks that compound, year after year.

3. The estate has moved from technical debt to strategic debt

Aging infrastructure no longer just slows down IT; it slows down the business. It limits agility, restricts transformation, and reduces competitiveness. CEOs may not use the words “technical debt,” but they understand when the organization is weighed down by the past.

How CIOs should reframe the conversation

To build trust and alignment, CIOs need to take ownership of this conversation rather than waiting for disruption to force it, and CEOs want three things from them. 

They want issues surfaced early and directly, with no surprises. CIOs need to lead with transparency.

Second is proactivity and the confidence to embrace change, make bold strategic calls, and recognize that even small fixes can have outsized impact, especially in an AI-driven environment.

And third is practicality. CEOs aren’t interested in “new toys,” but in well-evidenced, sensible solutions that reduce risk and address problems decisively when they arise. 

Above all, they want CIOs to think long term, planning infrastructure over the next decade rather than the next budget cycle and moving beyond an “if it isn’t broken” mindset. 

The moment for this conversation is now

Most enterprises are at an inflection point. Modernize the estate to unlock AI-driven advantage or carry forward a legacy footprint that cannot support the ambitions the CEO now expects the CIO to deliver. The CIO who leads this conversation will be seen as a true strategic partner.  

Explore all six crucial conversations

The IT estate is only one of six crucial conversations CIOs need to master with their CEO. To dive deeper into the rest – cost, risk, innovation, people and measurement – read the full Crucial Conversations report now. 

US federal software reform bill aims to strengthen software management controls

4 December 2025 at 11:57

Software management struggles that have pained enterprises for decades cause the same anguish to government agencies, and a bill making its way through the US House of Representatives to strengthen controls around government software management holds lessons for enterprises too.

The Strengthening Agency Management and Oversight of Software Assets (SAMOSA) bill, H.R. 5457, received unanimous approval from a key US House of Representative committee, the Committee on Oversight and Government Reform, on Tuesday.

SAMOSA is mostly focused on trying to fix “software asset management deficiencies” as well as requiring more “automation of software license management processes and incorporation of discovery tools,” issues that enterprises also have to deal with.

In addition, it requires anyone involved in software acquisition and development to be trained in the agency’s policies and, more usefully, in negotiation of contract terms, especially those that put restrictions on software deployment and use.

This training could also be quite useful for enterprise IT operations. It would teach “negotiating options” and specifically the “differences between acquiring commercial software products and services and acquiring or building custom software and determining the costs of different types of licenses and options for adjusting licenses to meet increasing or decreasing demand.”

The mandated training would also include tactics for measuring “actual software usage via analytics that can identify inefficiencies to assist in rationalizing software spending” along with methods to “support interoperable capabilities between software.”

Outlawing shadow IT

The bill also attempts to rein in shadow IT by “restricting the ability of a bureau, program, component, or operational entity within the agency to acquire, use, develop, or otherwise leverage any software entitlement without the approval of the Chief Information Officer of the agency.” But there are no details about how such a rule would be enforced.

It would require agencies “to provide an estimate of the costs to move toward more enterprise, open-source, or other licenses that do not restrict the use of software by the agency, and the projected cost savings, efficiency measures, and improvements to agency performance throughout the total software lifecycle.” But the hiccup is that benefits will only materialize if technology vendors change their ways, especially in terms of transparency.

However, analysts and consultants are skeptical that such changes are likely to happen.

CIOs could be punished

Yvette Schmitter, a former Price Waterhouse Coopers principal who is now CEO of IT consulting firm Fusion Collective, was especially pessimistic about what would happen if enterprises tried to follow the bill’s rules.

“If the bill were to become law, it would set enterprise CIOs up for failure,” she said. “The bill doubles down on the permission theater model, requiring CIO approval for every software acquisition while providing zero framework for the thousands of generative AI tools employees are already using without permission.”

She noted that although the bill mandates comprehensive assessments of “software paid for, in use, or deployed,” it neglects critical facets of today’s AI software landscape. “It never defines how you access an AI agent that writes its own code, a foundation model trained on proprietary data, or an API that charges per token instead of per seat,” she said. “Instead of oversight, the bill would unlock chaos, potentially creating a compliance framework where CIOs could be punished for buying too many seats for a software tool, but face zero accountability for safely, properly, and ethically deploying AI systems.”

Schmitter added: “The bill is currently written for the 2015 IT landscape and assumes that our current AI systems come with instruction manuals and compliance frameworks, which they obviously do not.”

She also pointed out that the government seems to be working at cross-purposes. “The H.R. 5457 bill is absurd,” she said. “Congress is essentially mandating 18-month software license inventories while the White House is simultaneously launching the Genesis Mission executive order for AI that will spin up foundation models across federal agencies in the next nine months. Both of these moves are treating software as a cost center and AI as a strategic weapon, without recognizing that AI systems are software.”

Scott Bickley, advisory fellow at Info-Tech Research Group, was also unimpressed with the bill. “It is a sad, sad day when the US Federal government requires a literal Act of Congress to mandate the Software Asset Management (SAM) behaviors that should be in place across every agency already,” Bickley said. “One can go review the [Office of Inspector General] reports for various government agencies, and it is clear to see that the bureaucracy has stifled all attempts, assuming there were attempts, at reining in the beast of software sprawl that exists today.”

Right goal, but toothless

Bickley said that the US government is in dire need of better software management, but that this bill, even if it was eventually signed into law, would be unlikely to deliver any meaningful reforms. 

“This also presumes the federal government actually negotiates good deals for its software. It unequivocally does not. Never has there been a larger customer that gets worse pricing and commercial terms than the [US] federal government,” Bickley said. “At best, in the short term, this bill will further enrich consultants, as the people running IT for these agencies do not have the expertise, tooling, or knowledge of software/subscription licensing and IP to make headway on their own.”

On the bright side, Bickley said the goal of the bill is the right one, but the fact that the legislation didn’t deliver or even call for more funding makes it toothless. “The bill is noble in its intent. But the fact that it requires a host of mandatory reporting, [Government Accountability Office] oversight, and actions related to inventory and overall [software bill of materials] rationalization with no new budget authorization is a pipe dream at best,” he said. 

Sanchit Vir Gogia, the chief analyst at Greyhound Research, was more optimistic, saying that the bill would change the law in a way that should have happened long ago.

“[It] corrects a long-standing oversight in federal technology management. Agencies are currently spending close to $33 billion every year on software. Yet most lack a basic understanding of what software they own, what is being used, or where overlap exists. This confusion has been confirmed by the Government Accountability Office, which reported that nine of the largest agencies cannot identify their most-used or highest-cost software,” Gogia said. “Audit reports from NASA and the Environmental Protection Agency found millions of dollars wasted on licenses that were never activated or tracked. This legislation is designed to stop such inefficiencies by requiring agencies to catalogue their software, review all contracts, and build plans to eliminate unused or duplicate tools.”

Lacks operational realism

Gogia also argued, “the added pressure of transparency may also lead software providers to rethink their pricing and make it easier for agencies to adjust contracts in response to actual usage.” If that happens, it would likely trickle into greater transparency for enterprise IT operations. 

Zahra Timsah, co-founder and CEO of i-GENTIC AI, applauded the intent of the bill, while raising logistical concerns about whether much would ultimately change even if it ultimately became law.

“The language finally forces agencies to quantify waste and technical fragmentation instead of talking about it in generalities. The section restricting bureaus from buying software without CIO approval is also a smart, direct hit on shadow IT. What’s missing is operational realism,” Timsah said. “The bill gives agencies a huge mandate with no funding, no capacity planning, and no clear methodology. You can’t ask for full-stack interoperability scoring and lifecycle TCO analysis without giving CIOs the tools or budget to produce it. My concern is that agencies default to oversized consulting reports that check the box without actually changing anything.”

Timsah said that the bill “is going to be very difficult to implement and to measure. How do you measure it is being followed?” She added that agencies will parrot the bill’s wording and then try to hire people to manage the process. “It’s just going to be for optic’s sake.”

The year ahead: What will become the 3 pillars of trust in an AI-first world?

4 December 2025 at 11:34

Today, the conversation in every boardroom is most likely centered on a single, transformative force: artificial intelligence (AI). Many see it as the engine for unprecedented growth, efficiency, and innovation. And, while this belief is justifiable, the entire revolution is being built on a fragile foundation of trust — an already fragile ground that is about to shift even further.

As AI systems begin to manage supply chains, deploy code, and execute financial transactions, the nature of risk changes entirely. The primary threat becomes the catastrophic cost of disruption to the intelligent systems that form the central nervous system of modern business.

To harness AI’s promise while mitigating its existential risks, we already know that leaders must move beyond a defensive security posture. To be effective, leaders must also fundamentally shift how they view security as a whole. They must view it as the foundation that innovation is built on, not as a barrier to progress. To do this, we, as a collective, must build a proactive strategy based on three core pillars of trust.

1. Engineering for trust

Trust cannot be an afterthought; it must be an engineering outcome. In the past, security was often a gate that slowed progress. Today, that model is inverted. A modern, unified security platform with trust built in by design now serves as a powerful strategic accelerator.

Automated security, when treated as a native component of the AI development lifecycle, eliminates the traditional brakes on progress. This enables our teams to innovate and deploy new models with the speed and confidence that delivers a direct, quantifiable competitive advantage. This transition from a reactive posture to one that ensures innovation velocity is key.

The “engineering for trust” approach also allows us to address a silent liability plaguing many organizations: decades of accumulated security debt. A patchwork of disconnected point products creates a complex and vulnerable attack surface, a problem now amplified by the cloud. Our exclusive internal research found that a majority of cloud databases related to AI development are not properly secured, lacking basic encryption or access controls.

Moving to a unified, trustworthy platform is akin to refinancing this debt — a solution that any board member would be amenable to. This type of platform simplifies operations, reduces long-term risk, and frees up our most valuable resources to focus on growth instead of just defense.

2. Cultivating cultures of trust

A single human error can undermine even the most perfectly engineered system. While technology provides the foundation, a vigilant and security-conscious culture forms the crucial human layer of the trust stack.

In an era of AI-powered phishing and sophisticated social engineering, every employee must become a steward of their organization’s security. This challenge is magnified by the rise of shadow AI. Our latest research on SaaS risks reveals that the use of unsanctioned third-party AI tools in the enterprise has skyrocketed, creating a massive blind spot where sensitive corporate data is regularly fed into untrusted models. That is why this pillar demands more than annual training videos. It requires a deep-seated culture of awareness where people are empowered to question anomalies and act as the first line of defense.

The value of this culture extends far beyond risk mitigation. A strong culture provides the ethical guardrails that ensure AI is used responsibly, protecting the brand and maintaining customer confidence that is so difficult to earn and so easy to lose. Its essential, human-driven process protects the organization from the inside out.

3. Governing for trust

The speed and scale of modern AI demand a new governance model built on two key principles: unwavering human control and radical industry-wide cooperation.

First, we must design systems that guarantee human oversight. Robust, human-in-the-loop governance is the ultimate safeguard against the catastrophic business disruption that autonomous systems could otherwise trigger. It is the board-level guarantee that our most valuable tools remain under our command, operating as intended.

Second, we must recognize that we cannot face this new threat landscape alone. AI-powered attacks are an ecosystem-wide problem that demands an ecosystem-wide defense. Sharing threat intelligence and best practices across companies and industries is a core business necessity for our collective survival and stability.

Trust as the ultimate ROI

To lead in the age of AI, our strategy must be clear. We need well-engineered systems that accelerate the business, a vigilant culture that protects it, and a robust governance that ensures its resilience. The goal of a modern security strategy has fundamentally changed, shifting from merely preventing incidents to actively creating and protecting value.

In the AI-first world, thriving organizations will understand that trust is the most valuable asset on their balance sheet and the ultimate driver of their success.

Curious what else Ben has to say? Check out his other articles on Perspectives.

Building tech leaders who think like CEOs (and deliver like operators)

4 December 2025 at 10:19

So your newly promoted CTO walks into their first executive meeting, armed with deep technical expertise and genuine enthusiasm for transformation. Six months later, they’re frustrated, your digital initiatives have stalled and your board is questioning the technology leadership strategy.

This isn’t a story about hiring the wrong person. It’s a story about building the wrong environment.

Here’s the truth your consultants won’t share: When technical leaders fail, it’s rarely a failure of intelligence. It’s a failure of integration.

Charles Sims notes this in his analysis of C-suite dynamics, “If you’re seated in the ‘big chair,’ you can’t expect people to intuit where they need to go. You need to build the bridge.”

The organizations winning the transformation race aren’t just hiring better CTOs; they’re creating fundamentally different conditions for technology leadership to thrive.

The hidden architecture of failure

Before we dive into solutions, let’s diagnose what’s actually broken.

The problem isn’t individual competence, it’s institutional design.

Most C-suite structures were established when technology was viewed as a cost center, rather than a competitive weapon. The processes, meeting rhythms and decision-making frameworks assume technology comes after strategy, not during it.

This creates what I call the integration gap, the space between where technology leaders sit and where they need to be to drive real transformation.

Deloitte research on resilient technology functions reveals a telling insight: High-performing “tech vanguard” businesses fundamentally differ in how they structure technology leadership.

As Khalid Kark and Anh Nguyen Phillips point out, these organizations embrace “joint accountability” and “establish sensing mechanisms that help anticipate business change.”

Translation: They don’t just include technology in business strategy, they integrate it.

The strategic exclusion problem

Here’s the most expensive mistake organizations make: bringing technology leaders into strategy validation, not strategy formation.

I’ve watched this pattern across dozens of transformations. The business leadership team spends months crafting the digital strategy. They debate market positioning, customer experience and competitive responses. Then, in the final act, they bring in the CTO to confirm technical feasibility.

This isn’t collaboration, it’s a recipe for execution failure.

CIO advisor Isaac Sacolick sums it up nicely, “What the risk here for CIOs is to get something out there on paper and start communicating. Letting your business partners know that you’re going to be the center point of putting a strategy together.

“Being able to do blue sky planning with business leaders, with technologists and data scientists on a very frequent basis to say, ‘is our strategy aligned or do we need a pivot’ or do we need to add I think that’s really the goal for a CIO now is to continually do that over the course of how this technology is changing.”

When technologists inherit fully formed strategies, they inherit the constraints, assumptions and blind spots of non-technical decision-making. The result? Strategies that sound compelling in PowerPoint but break down in reality.

The integration solution: As Sims emphasizes, successful businesses bring technology leaders in “when the goals are still being shaped.” Technology leaders become co-architects of strategy, not just implementers of it.

The translation challenge

Every business talks about wanting CTOs who can “translate technical complexity into business value.”

But most create conditions that make effective translation impossible.

The problem isn’t that technology leaders can’t communicate. It’s that business leaders structure every interaction to discourage strategic thinking. Fifteen-minute slots for infrastructure decisions. “High-level only” constraints on technical briefings. Interruptions when discussions get into architectural details.

Sims captures the real need perfectly: “Ask them to explain how tech can enable outcomes, not just avoid outages.” But enabling outcomes requires time, context and genuine dialogue — not rapid-fire status updates.

The integration solution: Create forums for substantive technical dialogue. Allocate time for technology leaders to educate business counterparts on possibilities, constraints and trade-offs.

The four pillars of technology leadership integration

The rebel leaders I’ve studied don’t just talk about integration, they systematically engineer it. Here are the four pillars that separate transformation winners from digital theater performers.

Pillar one: Strategic co-creation

Instead of: Bringing technology leaders in for feasibility validation.

Rebels: Include them in strategic formation from day one.

The breakthrough insight is simple: Technology constraints and possibilities should shape strategy, not just constrain it. When technologists participate in strategic formation, they help identify opportunities that pure business thinking might miss.

Actionable implementation:

  • Include your CTO in quarterly business reviews, not just technology reviews
  • Require technology input before major strategic initiatives get funded
  • Create joint business-technology planning sessions for all transformation efforts
  • Give technology leaders access to the same market intelligence and customer feedback as other executives

Pillar two: Outcome-driven accountability

Instead of: Asking for deliverables and timelines.

Rebels: Define success in business outcomes and measure accordingly.

This shift eliminates the translation problem entirely. When success is defined in business terms from the beginning, technology leaders naturally think about impact, not just implementation.

The Deloitte study talks about “value-based investments” aligned with “iterative Agile sprints.” But the real innovation isn’t methodological, it’s definitional. Success gets measured by business value delivered, not features completed.

Actionable implementation:

  • Replace project status meetings with outcome review sessions
  • Tie technology leader compensation to business metrics, not just technical ones
  • Create shared dashboards that track business impact of technology initiatives
  • Require business case updates, not just project updates

Pillar three: Information symmetry

Instead of: Functional hierarchy with information silos.

Rebels: Ensure technology leaders have the same strategic context as business leaders.

Sims makes a crucial point: “Technology touches every department. The org chart should reflect that.” But organizational design goes beyond reporting structures; it’s about information flow and decision rights.

The Deloitte research highlights the need for “sensing mechanisms that help anticipate business change.” But sensing requires access to information, not just responsibility for reaction.

Actionable implementation:

  • Include technology leaders in customer advisory boards and market research reviews
  • Share competitive intelligence and industry analysis with the entire C-suite, not just business functions
  • Create cross-functional intelligence-sharing sessions where every leader contributes market insights
  • Ensure technology leaders participate in customer meetings and strategic partnerships

Pillar four: Translation excellence

Instead of: Expecting natural translation ability.

Rebels: Systematically develop two-way translation competence.

Here’s where most organizations get it backwards. They expect CTOs to be great translators but provide no development, feedback or support for this critical skill.

As Sims notes, “The best CTOs turn complexity into clarity. They make everyone around them smarter. That’s the leadership skill we should be measuring.”

But translation is a two-way street. Business leaders also need to develop competence in asking strategic questions that unlock technological insight.

Actionable implementation:

  • Create monthly translation labs where technology leaders practice explaining complex concepts to different audiences
  • Train business leaders to ask better questions: “What are the trade-offs?” instead of “Is this feasible?”
  • Establish technology education sessions for non-technical executives
  • Reward and recognize technology leaders who effectively educate their peers

Better leadership means faster business

When you get technology leadership integration right, the impact extends far beyond individual performance. You create what the Deloitte research calls enterprise agility: the ability to “nimbly strategize and operate” in response to constant change.

The data reveals so much: businesses with integrated technology leadership outperform peers across every meaningful metric. Revenue growth, profit margins, customer satisfaction, employee engagement and market share all improve when business and technology leadership truly collaborate.

But the most significant impact might be speed. Integrated organizations move faster because they eliminate the handoff delays, translation loops and rework cycles that plague siloed structures.

The competitive reality

While you’re optimizing technology leadership integration, your competitors are making a choice. Some will continue the old patterns: hiring smart technologists, giving them business requirements and wondering why transformation is hard.

Others will join the integration revolution. They’ll create conditions where technology leaders thrive. They’ll build strategic collaboration into their organizational DNA. They’ll accelerate past competitors while others struggle with digital theater.

The study reveals that tech vanguard organizations are already pulling away from baseline performers. The gap isn’t just technical: it’s structural, cultural and strategic.

Ready to ramp up?

The path forward isn’t about your next technology hire, it’s about the environment you create for technology leadership to succeed.

Week one: Audit your current integration points. Where does your CTO participate in strategic decision-making? Where are they excluded? Map the information flows and decision rights.

Month one: Redesign your leadership meeting rhythms. Include technology leaders in strategic formation, not just implementation planning. Create forums for substantive business-technology dialogue.

Month two: Implement outcome-based accountability. Replace deliverable tracking with business impact measurement. Align technology leader success metrics with business results.

Month three: Launch translation competence development. Create systematic programs for both business-to-technology and technology-to-business communication improvement.

Month six: Measure integration velocity. How quickly do business insights flow into technology decisions? How rapidly do technological possibilities inform business strategy?

The businesses that systematically build technology leadership integration won’t just transform their trajectory; they’ll transform their markets. They’ll set the pace while competitors struggle to keep up.

The choice is yours: Continue with traditional technology leadership models or build the integration capabilities that drive real transformation.

The rebels are already deciding. What about you?

This article is published as part of the Foundry Expert Contributor Network.
Want to join?

Why CIOs must reimagine ERP as the enterprise’s composable backbone

4 December 2025 at 09:20

In my experience leading ERP modernization projects and collaborating with IT and business executives, I’ve learned that technology alone rarely determines success, but mindset and architecture do. Gartner reports, “By 2027, more than 70% of recently implemented ERP initiatives will fail to fully meet their original business case goals.” ERP success now requires a fundamentally different architecture.

For decades, ERP systems have been the core of enterprise operations: managing finance, supply chain, manufacturing, HR and more. The same systems that once promised control and integration are now stifling flexibility, slowing innovation and piling up technical debt.

From what I’ve observed across multiple ERP programs, the problem isn’t ERP itself, but rather, it’s how we’ve come to think about it. Many organizations still treat ERP purely as a system of record, missing the broader opportunity in front of them.

The next era of business agility will be defined by ERP as a composable platform: modular, data-centric, cloud-native and powered by AI. In many of the organizations I’ve worked with, technology leaders aren’t debating whether to modernize the core. Instead, they’re focused on how to do it without stalling the business.

Forbes captures the shift succinctly: “it is anticipated that 75% of global businesses will begin replacing traditional monolithic ERP systems with modular solutions — driven by the need for enhanced flexibility and scalability in business operations.” This highlights ERP’s evolution from monolithic legacy suites to an adaptive, innovation-driven platform.

Those who embrace this shift will make ERP an enabler of innovation. Those who don’t will watch their core systems become their biggest bottleneck and stay held back.

From monoliths to modular backbones

In the 1990s and 2000s, ERP meant one vendor, one codebase and one massive implementation project touching every corner of the business. Companies spent millions customizing software to fit every process nuance.

I saw the next chapter unfold with the cloud era. Companies such as SAP, Oracle, Microsoft and Infor transitioned their portfolios to SaaS, while a wave of startups emerged with modular, industry-focused ERP platforms. APIs and services finally promised a system that could evolve with the business.

In one transformation I supported, our biggest turning point came when we stopped treating ERP as a single implementation. We began decomposing capabilities into modules that business teams could own and evolve independently.

But for many enterprises, that promise never fully materialized. The issue isn’t the technology anymore, but the mindset. In many organizations, ERP is still viewed as a finished installation rather than a living platform meant to grow and adapt.

The cost of the old mindset

Legacy ERP thinking simply can’t keep up with today’s pace of change. The result is slower innovation, fragmented data and IT teams locked in perpetual catch-up mode. Organizations need architectures that change as fast as the business does.

LeanIX, citing Gartner research, highlights the advantage: “Organizations that have adopted a composable approach to IT are 80% faster in new-feature implementation, particularly when using what Gartner defines as composable ERP platforms,” demonstrating the performance gap between modular ERP and traditional monolithic systems.

I’ve seen legacy ERP thinking carry a high price tag in real projects:

  • Inflexibility: Business models evolve faster than software cycles. Traditional ERP can’t keep up.
  • Over-customization: Years of bespoke code make upgrades risky and expensive.
  • Data fragmentation: Multiple ERP instances and disconnected modules create inconsistent data and unreliable analytics.
  • User frustration: Outdated interfaces drive workarounds and disengagement.
  • High total cost of ownership: Maintenance and upgrades consume budgets that should fund innovation.

Enter the composable ERP

The emerging composable ERP model breaks this monolith apart. Gartner defines it as an architecture where enterprise applications are assembled from modular building blocks, connected through APIs and unified by a data fabric.

As LeanIX explains, “Composable ERP, built on modular and interoperable components, allows organizations to respond faster to change by assembling capabilities as needed rather than relying on a rigid, monolithic suite,” illustrating the transition from static ERP systems to a dynamic, adaptable business platform.

Having worked on both sides — custom development and packaged ERP — I’ve learned that the real power of composability lies in how easily teams can assemble, not just integrate, capabilities. Rather than seeing ERP as a single suite, think of it as the system that enables how an enterprise operates. The core processes — finance, supply chain, manufacturing, HR — are what make up the base. Modular features such as AI forecasting, customer analytics and sustainability tracking can plug in dynamically as the business evolves.

This approach enables organizations to:

  • Mix and match modules from different vendors or in-house teams.
  • Integrate best-of-breed cloud apps through standard APIs instead of brittle custom code.
  • Leverage AI for automation, insights and predictive decisions.
  • Deliver persona-based experiences tailored to each user’s role.

Personas: The human face of composable ERP

Traditional ERP treated every user the same, in which there would be one interface, hundreds of menus, endless forms. Composable ERP flips that script with persona-based design, built around what each role needs to accomplish.

  • CFOs see real-time financial health across entities with AI-driven scenario modeling.
  • Supply chain leaders monitor live demand signals, supplier performance and sustainability metrics.
  • Plant managers track IoT-enabled equipment, predictive maintenance and production KPIs.
  • Sales and service teams access operational data seamlessly without switching systems.

From my experience, when ERP is designed around real personas rather than generic transactions, adoption rises and decisions happen faster.

Challenges and pitfalls

These are not theoretical issues; they’re the practical challenges I see IT and business teams grappling with every day.

  • Data governance: Without a unified data strategy, modularity turns to chaos.
  • Integration complexity: APIs require discipline for versioning, authentication, semantic alignment.
  • Vendor lock-in: Even open platforms can create subtle dependencies.
  • Change management: Employees need support and training to unlearn old habits.
  • Security: A more connected system means a larger attack surface. Zero-trust security is essential.

True success demands leadership that balances technical depth with organizational empathy.

The CIO’s new playbook

Through years of ERP work and collaboration between business and IT teams, I’ve realized that the biggest hurdle to ERP success is the belief that ERP is a fixed system instead of a constantly evolving platform for innovation.

This shift isn’t about tools, but rather it’s about redefining the ERP’s role in the business. McKinsey reinforces this reality, stating, “Modernizing the ERP core is not just a technology upgrade — it is a business transformation that enables new capabilities across the enterprise.” It’s a shift that calls for a fundamentally different playbook, especially for CIOs leading modernization efforts.

  1. Start with the business architecture, not the software. Define how you want your enterprise to operate, then design ERP capabilities to fit.
  2. Build a unified data fabric. A composable ERP lives or dies by consistent, high-quality data.
  3. Adopt modular thinking incrementally. Start small by piloting a few modules, prove the value, then scale.
  4. Empower fusion teams. Blend IT, operations and business experts into agile squads that compose solutions quickly.
  5. Measure success by outcomes, not go-lives. The goal is agility and resilience and not a single launch date.
  6. Push vendors for openness. Demand published APIs and true interoperability, not proprietary cloud labels.

Oracle reinforces this imperative: “Companies need to move toward a portfolio that is more adaptable to business change, with composable applications that can be assembled, reassembled and extended,” highlighting flexibility as a core selection criterion.

Reframe ERP as an innovation platform. Encourage experimentation with low-code workflows, analytics and AI copilots.

Looking ahead: When ERP becomes invisible

In a few years, we might not even use the term ERP. Like CRM’s evolution into customer experience platforms, ERP will fade into the background, becoming the invisible digital backbone of the enterprise.

I’ve watched ERP evolve from on-premises to cloud to AI-driven platforms. AI will soon handle transactions and workflows behind the scenes, while employees interact through conversational interfaces and embedded analytics. Instead of logging into systems, they’ll simply request outcomes — and the composable ERP fabric will dynamically orchestrate everything required to deliver them.

That future belongs to organizations rethinking ERP today. This isn’t just another upgrade cycle — it’s a redefinition of how enterprises operate.

From record-keeping to value creation

ERP was once about efficiency — tracking inventory, closing books, enforcing process discipline. Today, it’s about resilience and innovation. From my own journey across multiple ERP programs, I’ve seen that the CIO’s challenge isn’t just keeping systems running, but also architecting agility into how the enterprise operates.

Composable ERP, which is built on cloud, AI and human-centered design, is the blueprint. It turns ERP from a system of record into a system of innovation that evolves as fast as the market around it.

The opportunity is clear: Lead the transformation or risk maintaining yesterday’s architecture while others design tomorrow’s enterprise.

This article is published as part of the Foundry Expert Contributor Network.
Want to join?

LLM゚ヌゞェントずRAGの実務的な組み合わせ方──ドキュメント怜玢から“自埋アシスタント”ぞ

4 December 2025 at 07:19

なぜRAGだけでは「胜動的に動かない」のか

RAGの基本的な仕組みは比范的シンプルです。ナヌザヌからの質問を意味ベクトルに倉換し、その近くに䜍眮する瀟内ドキュメントやナレッゞ蚘事を怜玢し、その内容をもずにLLMが回答を生成したす。このアプロヌチにより、モデルが孊習しおいない最新の情報や䌁業固有のルヌルに基づいた回答が可胜になり、埓来のFAQボットよりも柔軟な応答が実珟したす。

しかし、RAGはあくたで「情報を取り出しおくる」胜力に特化した仕組みです。ナヌザヌから明瀺的な質問が来ない限り動き出さず、「次に䜕をすべきか」「どんな資料を䜜るべきか」ずいったタスクレベルの提案は自発的には行いたせん。たた、倚くの堎合、RAGシステムはチャットむンタヌフェヌスの内偎に閉じおおり、瀟内のワヌクフロヌや倖郚ツヌルずの連携は限定的です。その結果、䟿利ではあるものの「怜玢窓が少し賢くなった」皋床のむンパクトに留たり、業務党䜓の生産性を倧きく倉えるずころたでは到達しおいないケヌスが目立ちたす。

この状況を打砎するためには、「情報を探しお答える」から䞀歩進んで、「目的達成のために情報を取りに行き、䜿いながらタスクを進める」存圚ずしおRAGを組み蟌む必芁がありたす。ここで鍵になるのが、LLM゚ヌゞェントずの統合です。

RAGをツヌルずしお扱う゚ヌゞェントの蚭蚈

RAGず゚ヌゞェントを統合する最も自然な方法は、RAGを゚ヌゞェントから呌び出す「䞀぀のツヌル」ずしお扱うこずです。゚ヌゞェントは、タスクを遂行する䞭で「瀟内ドキュメントに関連情報がないか調べる必芁がある」ず刀断するず、RAGツヌルを呌び出したす。RAGは質問文を受け取り、関連資料ずその芁玄を返したす。゚ヌゞェントは、その結果を読み解きながら蚈画を曎新し、次のアクションを決めおいきたす。

䟋えば、新補品の提案曞を䜜るタスクを考えおみたす。ナヌザヌが゚ヌゞェントに察しお「〇〇業界向けの提案曞の叩き台を䜜っお」ず指瀺するず、゚ヌゞェントはたずRAGを䜿っお、過去の提案曞や関連する技術資料、業界レポヌトを怜玢したす。その内容を芁玄し、今回の提案に䜿えそうな芁玠を敎理し、提案曞の構成案を䜜りたす。続いお、各セクションのドラフトを曞きながら、必芁に応じおRAGで远加情報を取りに行きたす。最終的には、提案曞の骚子ず本文ドラフト、参考資料の䞀芧たでをたずめお提瀺するこずができたす。

この蚭蚈のポむントは、RAGが単独でナヌザヌず察話するのではなく、゚ヌゞェントの「目」ずしお機胜しおいるこずです。ナヌザヌのゎヌルず党䜓の流れを理解しおいるのぱヌゞェント偎であり、RAGは必芁なずきに呌び出される知識リ゜ヌスずいう䜍眮づけになりたす。こうするこずで、RAGが持぀「情報ぞのアクセス胜力」ず、゚ヌゞェントが持぀「タスク遂行胜力」が自然に統合されたす。

ワヌクフロヌ党䜓の䞭でRAG゚ヌゞェントを䜍眮づける

RAGず゚ヌゞェントを組み合わせるずき、個別のチャット䜓隓だけに泚目するのではなく、業務ワヌクフロヌ党䜓の䞭でどのステップを眮き換えるのかを蚭蚈する芖点が重芁です。たずえば、瀟内の皟議曞䜜成プロセスを考えおみるず、関連芏皋や過去の皟議曞の確認、文面のドラフト䜜成、必芁な資料の添付、関係者ぞの説明資料の準備など、倚くのステップがありたす。

この䞀連の流れの䞭で、RAGは芏皋や過去事䟋の怜玢に匷みがありたす。䞀方、゚ヌゞェントは「どの項目をどう埋めるべきか」「どの資料を添付すべきか」ずいったタスクの流れをコントロヌルするこずができたす。したがっお、理想的には、皟議システムの画面そのものに゚ヌゞェントを組み蟌み、ナヌザヌが皟議を新芏䜜成するず、゚ヌゞェントが必芁項目の入力をガむドしながら、裏偎でRAGを䜿っお参考情報を探し、フィヌルドごずに候補を埋めおいくような䜓隓が望たしいでしょう。

たた、ワヌクフロヌ゚ンゞンず連携させるこずで、゚ヌゞェントが単なる文曞䜜成支揎を超えお、次のステップのトリガヌ圹を担うこずもできたす。たずえば、ある条件を満たした皟議が提出されたら、゚ヌゞェントが自動的に関連郚門のレビュヌ担圓者ぞ説明メモを送付し、必芁に応じお質疑応答のサポヌトも行う、ずいった具合です。ここでもRAGは、過去の類䌌案件やQ&Aログの怜玢に䜿われ、゚ヌゞェントは察話ずタスク管理の䞡方を叞る存圚ずしお動きたす。

もちろん、こうした統合には泚意点もありたす。RAGが返す情報は垞に正確ずは限らず、叀い資料や誀った文曞が含たれる可胜性もありたす。そのため、゚ヌゞェント偎で「情報の新しさ」や「瀟内での信頌床」を刀断するロゞックを組み蟌んだり、人間のレビュヌを挟むポむントを明瀺したりするこずが重芁です。たた、RAGにどの範囲の瀟内文曞を読み蟌たせるかずいう暩限蚭蚈も、セキュリティず利䟿性のバランスを巊右したす。

RAGず゚ヌゞェントの組み合わせは、単なる技術的なトリックではなく、䌁業内の情報フロヌを再蚭蚈するための匷力なレバヌです。ドキュメント怜玢の延長線䞊にずどたらず、「この組織で人が日々行っおいる情報収集ず意思決定プロセスを、どこたで゚ヌゞェントに肩代わりさせるか」を考えるこずが、真の䟡倀を匕き出す鍵になりたす。

❌
❌