So you keep hearing about how Home Assistant is the ultimate smart home platform, but you still feel lost. Here are some burning questions you might need answered before you can take the next step.
Setting up access to a headless Raspberry Pi is one of those tasks that should take a few minutes, but for some reason always seems to take much longer. The most common method is to configure Wi-Fi access and an SSH service on the Pi before starting it, which can go wrong in many different ways. This author, for example, recently spent a few hours failing to set up a headless Pi on a network secured with Protected EAP, and was eventually driven to using SSH over Bluetooth. This could thankfully soon be a thing of the past, as [Paul Oberosler] developed a package for SSH over USB, which is included in the latest versions of Raspberry Pi OS.
The idea behind rpi-usb-gadget is that a Raspberry Pi in gadget mode can be plugged into a host machine, which recognizes it as a network adapter. The Pi itself is presented as a host on that network, and the host machine can then SSH into it. Additionally, using Internet Connection Sharing (ICS), the Pi can use the host machine’s internet access. Gadget mode can be enabled and configured from the Raspberry Pi Imager. Setting up ICS is less plug-and-play, since an extra driver needs to be installed on Windows machines. Enabling gadget mode only lets the selected USB port work as a power input and USB network port, not as a host port for other peripherals.
An older way to get USB terminal access is using OTG mode, which we’ve seen used to simplify the configuration of a Pi as a simultaneous AP and client. If you want to set up headless access to Raspberry Pi desktop, we have a guide for that.
Thanks to [Gregg Levine] for the tip!
Raspberry Pis are popular because they're small, don't consume much power, and are extremely versatile. However, the default method of booting one—from the microSD card—is best avoided if possible. Here's how you can do it.
Raspberry Pi devices are popular among retro enthusiasts looking to emulate old computers and consoles, but this usually only goes as far as software. What you might not have considered is that the Raspberry Pi can also play a role in reviving old hardware.
New year, new Startup Radar.
We’re back with our regular spotlight on early stage startups sprouting up in the Seattle region. For this edition, we’re featuring Axel AI, DrunR, Eluum, and profileAPI.
Read on for brief descriptions of each company — along with pitch assessments from “Mean VC,” a GPT-powered critic offering a mix of encouragement and constructive criticism.
Check out past Startup Radar posts here, and email me at taylor@geekwire.com to flag other companies and startup news.
Founded: 2025
The business: A self-described “reasoning layer” for retail media sales teams that aims to translate messy data into commercial narratives and proposals. The idea is to help sales teams spend less time on manual analysis and preparation. The bootstrapped company officially launched its MVP at CES and NRF 2026 earlier this month.
Leadership: CEO and co-founder Bobby Figueroa previously founded Gradient, another Seattle-based commerce insights company that was acquired by Criteo. He was also an exec at Amazon. Axel’s leadership and advisory team includes former sales and advertising leaders at Amazon, Google, and Microsoft.
Mean VC: “You’re targeting a real friction point — sales teams juggling fragmented data with limited time to craft a compelling narrative. The pedigree helps, but long-term success will hinge on whether your product drives actual revenue lift, not just cleaner decks. I’d focus on embedding directly into the sales team’s existing workflow — don’t make users open another tool, make yours the one that quietly does the heavy lifting behind the scenes.”
Founded: 2024
The business: A nutrition app that provides personalized guidance based on users’ goals and preferences, particularly while dining out or ordering food online. DrunR is running a closed beta in Seattle with restaurants and users, including people using GLP-1 medication. The startup is part of the WTIA Founder Cohort 13 program.
Leadership: Founder and CEO Yaya Ali is a financial analyst at Perkins Coie and previously worked for King County and Amazon. He also has food operations experience. David Greene, the company’s CTO, is a software engineer at Capital One and previously worked at Moody’s.
Mean VC: “The intersection of nutrition, personalization, and GLP-1s is timely — especially as eating habits shift alongside new weight-loss drugs. The challenge will be making the app feel essential day-to-day, not just ‘nice to have’ after a restaurant meal or clinic visit. I’d zero in on a high-frequency use case — something that keeps users opening the app daily, not just when they’re thinking about dinner.”
Founded: 2024
The business: A new take on social media with a product that helps people organize their personal memories, stories, and digital artifacts into one user-controlled system. Built on community-driven moderation and works across different platforms. The bootstrapped company is onboarding early users and plans to launch a MVP later this year.
Leadership: CEO and co-founder Bilkay Rose was a VP at tax software company Avalara and a director at Clearwire. Other co-founders include CTO Dale Rector, who spent three decades at Microsoft, and Jennifer Gianola, also a former exec at Avalara.
Mean VC: “The concept taps into a real emotional need — people are overwhelmed by digital clutter and increasingly skeptical of algorithm-driven feeds. The key will be showing how your platform earns daily use without relying on dopamine loops. I’d push to define a sharp use case first — memory curation is broad, so lead with one thing people urgently want to preserve, then expand once you’ve earned their trust.”
Founded: 2024
The business: A business data layer for developers building AI-native chat, copilot, and agentic tools for go-to-market. Its platform tracks more than 10,000 signals across more than 10 million companies and 500 million professionals. The company, which was previously a sales AI agent product called Truebase, has raised $2 million in funding.
Leadership: Founder and CEO Wissam Tabbara has sold two startups and spent more than six years at Microsoft in the 2000s.
Mean VC: “The shift from product to platform is smart — selling infrastructure to power GTM copilots has stronger upside than building another agent. But you’ll need to show that your data isn’t just broad, but relevant and timely enough to drive meaningful in-app decisions. I’d focus on becoming the plug-and-play GTM brain — make integration dead simple, and let other tools build magic on top of your stack.”
Key takeaways
PI, the native coin of the Pi network, has lost 1.6% of its value in the last 24 hours and is now trading above $0.18.
The bearish performance comes despite Pi Network announcing plans on Wednesday to boost the ecosystem, including a creator event, integration of the PI payments system into apps built on the network, and extended access to app creation.
The team revealed that the PI payments support is limited to Test-Pi, and new or non-migrated Pioneers can now deploy app iterations by watching ads instead of paying fees.
Furthermore, Pi Network believes that the ad-supported application building on Pi App Studio could reduce the financial burden of creating Pi applications.
In addition to that, retail demand continues to increase despite PI’s price decline over the past few days. Data obtained from PiScan shows that the users have removed 1.17 million PI tokens from CEXs over the past 48 hours.
The removal from central exchanges will decrease selling pressure on PI as the tokens are transferred to long-term wallets.
The PI/USDT 4-hour chart is bearish and efficient as Pi has lost 1.6% of its value in the last 24 hours. PI failed to maintain its rally above the $0.1919 support-turned-resistance level, marked by the October 11 low.
At press time, PI is trading at $0.1839. If the selloff continues, PI could retest the October 10 and January 19 lows at $0.1533 and $0.1502, respectively.
Technical indicators on the 4-hour chart suggest that the bears remain in control. The Relative Strength Index (RSI) is 40, below the neutral 50, while the Moving Average Convergence Divergence (MACD) extends below the signal line.
However, if the bulls regain control and PI closes its daily candle above $0.1919, it could further extend the rally, potentially targeting the December 19 high at $0.2177.
The post PI could slip below $0.17 despite payments update: Check forecast appeared first on CoinJournal.
Are you ready for a couple of challenging (and one more simple) Raspberry Pi projects? This weekend’s Raspberry Pi projects will put your hardware skills to the test with tasks like handling stepper motors and drivers, programming LEDs, and spinning up Docker containers.
Raspberry Pi is continuing its expansion into hardware accessories and components. In addition to SD cards and SSDs, you can now get an official Raspberry Pi Flash Drive. It's intended to be a boot drive for Pi boards, but you can use it like any other flash drive as well.
The ability to swap operating systems on a whim is one of the best things about single-board computers like the Raspberry Pi. You can write a new operating system to a memory card whenever you fancy, and keep a small collection of cards handy whenever you want to change.
Raspberry Pi boards have supported USB gadget mode for years, allowing you to connect a Pi to a computer with one cable for power and SSH/desktop access. Now, there's a much easier way to try it out, and it's built into Raspberry Pi OS.
API cybersecurity will be a ping pong ball, battered between the rackets of AI-assisted attackers and AI-assisted defenders.
The post Cyber Insights 2026: API Security – Harder to Secure, Impossible to Ignore appeared first on SecurityWeek.
Raspberry Pi single-board computers are awesome for so many reasons—they're low-power, extremely capable, and have small footprints. I love all those things about my Raspberry Pi, but my favorite feature is one that I don’t hear talked about nearly often enough.
Key takeaways
PI, the native coin of the Pi Network, is up 1% in the last 24 hours and is now trading at $1.91 per coin. The positive performance comes despite the broader cryptocurrency market recording losses in the last few hours.
According to PiScan, the reserves of centralized exchanges have decreased by 4.24 million PI tokens, indicating large withdrawals over the last 24 hours. The decline in exchange reserves reflects strong buying pressure, allowing PI to recover above $0.19.
The PI/USDT 4-hour chart is bearish and efficient despite the coin adding 1% to its value in the last 24 hours. At press time, PI is trading at $0.191, roughly 30% up from Monday’s low at $0.1502. The recovery aligns with the strong buying pressure and could push PI’s price higher in the near term.
The RSI of 33 means that PI is slowly escaping the oversold region as buyers step in. The MACD lines are still within the negative territory, indicating that the sellers have yet to fully relinquish control.
If the recovery continues and PI hits the $0.1919 resistance level, it could rally towards the $0.2060 psychological zone. An extended bullish run would allow PI hit the previous weekly high of $0.2116.
However, a daily candle close below $0.1919 could see PI give up some of its recent gains and retest the support levels at $0.1835 and $0.1632 in the near term.
The post PI rebounds above $0.19 despite selling pressure: Check forecast appeared first on CoinJournal.
Welcome back, aspiring forensics investigators.
In the previous article, we lifted the curtain on tracking technologies and showed how much information the internet collects from you. Many people still believe that privacy tools such as VPNs completely protect them, but as you are now learning, the story goes much deeper than that. Today we will explore what else is hiding behind the code. You will discover that even more information can be extracted from your device without your knowledge. And of course, we will also walk through ways to reduce these risks, because predictability creates patterns. Patterns can be tracked. And tracking means exposure.
Most people assume fingerprinting is only about what you see on the screen. However, browser fingerprinting reaches far beyond the visual world. It also includes non visual methods that silently measure the way your device processes audio or stores small website assets. These methods do not rely on cookies or user logins. They do not require permission prompts. They simply observe tiny differences in system behavior and convert them into unique identifiers.
A major example is AudioContext fingerprinting. This technique creates and analyzes audio signals that you never actually hear. Instead, the browser processes the sound internally using the Web Audio API. Meanwhile favicon based tracking abuses the way browsers cache the small icons you see in your tab bar. Together, these methods help trackers identify users even if visual fingerprints are blocked or randomized. These non visual fingerprints work extremely well alongside visual ones such as Canvas and WebGL. One type of fingerprint reveals how your graphics hardware behaves. Another reveals how your audio pipeline behaves. A third records caching behavior. When all of this is combined, the tracking system becomes far more resilient. It becomes very difficult to hide, because turning off one fingerprinting technology still leaves several others running in the background.
Everything occurs invisibly behind the web page. Meanwhile your device is revealing small but deeply personal technical traits about itself.
AudioContext fingerprinting is built on the Web Audio API. This is a feature that exists in modern browsers to support sound generation and manipulation. Developers normally use it to create music, sound effects, and audio visualizations. Trackers, however, discovered that it can also be used to uniquely identify devices.
Here is what happens behind the scenes. A website creates an AudioContext object. Inside this context, it often generates a simple sine wave using an OscillatorNode. The signal is then passed through a DynamicsCompressorNode. This compressor highlights tiny variations in how the audio is processed. Finally, the processed audio data is read, converted into numerical form, and hashed into an identifier.
The interesting part is where the uniqueness comes from. Audio hardware varies greatly. Different manufacturers like Realtek or Intel design chips differently. Audio drivers introduce their own behavior. Operating systems handle floating point math in slightly different ways. All of these variations influence the resulting signal, even when the exact same code is used. Two computers will nearly always produce slightly different waveform results.
Only specific privacy protections can interfere with this process. Some browsers randomize or block Web Audio output to prevent fingerprinting. Others standardize the audio result across users so that everyone looks the same. But if these protections are not in place, your system will keep producing the same recognizable audio fingerprint again and again.
You can actually test this yourself. There are demo websites that implement AudioContext fingerprinting.
Favicons are the small images you see in your browser tabs. They appear completely harmless. However, the way browsers cache them can be abused to create a tracking mechanism. The basic idea is simple. A server assigns a unique identifier to a user and encodes that identifier into a specific pattern of favicon requests. Because favicons are cached separately from normal website data, they can act as a form of persistent storage. When the user later returns, the server instructs the browser to request a large set of possible favicons. Icons that are already present in the cache do not trigger network requests, while missing icons do. By observing which requests occur and which do not, the server can reconstruct the original identifier.
This is clever because favicon caches have traditionally been treated differently from normal browser data. Clearing cookies or browsing history often does not remove favicon cache entries. In some older browser versions, favicon cache persistence even extended across incognito sessions.
There are limits. Trackers must maintain multiple unique icon routes, which requires server side management. Modern browsers have also taken steps to partition or isolate favicon caches per website, reducing the effectiveness of the method. Still, many legacy systems remain exposed, and clever implementations continue to find ways to abuse caching behavior.
Fingerprinting does not stop with visuals and audio. There are many additional identifiers that leak information about your device. Screen fingerprinting gathers details such as your screen resolution, usable workspace, color depth, pixel density, and zoom levels. These factors vary across laptops, desktops, tablets, and external monitors.
Font enumeration checks which fonts are installed on your system. This can be done by drawing hidden text elements and measuring their size. If the size changes, the font exists. The final list of available fonts can be surprisingly unique.
Speech synthesis fingerprinting queries the Web Speech API to discover which text to speech voices exist on your device. These are tied to language packs and operating system features.
The Battery Status API can reveal information about your battery capacity, charge state, and discharge behavior. This information itself is not very useful, but it helps illustrate how deep browser fingerprinting can go.
The website may also detect which Chrome plugins you use, making your anonymous identity even more traceable.
And this is still only part of the story. Browsers evolve quickly. New features create new opportunities for fingerprinting. So awareness is critical here.
When audio fingerprinting, favicon identifiers, Canvas, WebGL, and other methods are combined, they form what is often called a super fingerprint. This is a multi-layered identity constructed from many small technical signals. It becomes extremely difficult to change without replacing your entire hardware and software environment. This capability can be used for both legitimate analytics and harmful surveillance. Advertisers may track behavior across websites. Data brokers may build profiles over time. More dangerous actors may attempt to unmask users who believe they are anonymous.
Fortunately, there are tools that help reduce these risks. No defense is perfect. But layered protections can improve your privacy. For example, Tor standardizes many outputs, including audio behaviors and cache storage. But not everything, which means some things can expose you. Firefox includes settings such as privacy.resistFingerprinting that limit API details. Brave Browser randomizes or blocks fingerprinting attempts by default. Extensions such as CanvasBlocker and uBlock Origin also help reduce exposure, although they must be configured with care.
We encourage you to test your own exposure, experiment with privacy tools, and make conscious decisions about how and where you browse.
The key takeaway is not paranoia. Privacy tools do not eliminate fingerprinting, but defenses such as Tor, Brave, Firefox fingerprint-resistance, and well-configured extensions do reduce exposure. Understanding how non-visual fingerprints work allows you to make informed decisions instead of relying on assumptions. In modern browsing, privacy is not about hiding perfectly. It is about minimizing consistency and breaking long-term patterns.
Awareness matters. When you understand how you are being tracked, you’re far better equipped to protect your privacy.
Home Assistant is a Linux-based smart home operating system with a very low barrier to entry. In addition to being completely free, its minimum system specifications mean it works on most computers produced within the last 10 to 15 years (plus humble single-board computers).
Not all Raspberry Pi projects require a display, like home servers and DNS sinkholes. But not all projects are alike, and sometimes your choice of display can greatly influence the end result.
Recently, [Vinnie] aka [vinthewrench] moved from Oregon to Arkansas to start a farmstead. This is a style of farming that focuses not just on a profitable farm where produce is sold at market, but also on a homestead where much of one’s own food is grown on the farm as well. Like any farm, though, it’s extremely hard work that takes a tremendous amount of time. Automation and other technology can make a huge impact in these situations, and [Vinnie] is rolling out his own software stack to help with this on his farm.
He calls his project the Pi Internet of Things, or PioT, and as its name suggests is based around the Raspberry Pi. Since this will all be outdoors and exposed to the extremes of Arkansas weather, everything built under the auspices of this project prioritizes ruggedness, stability, and long-term support, all while avoiding any cloud service. The system also focuses on being able to ride through power outages. The server side, called piotserver, uses a REST API to give the user access to the automation systems through a web interface
[Vinnie] also goes into detail about why existing systems like Home Assistant and Open Sprinkler wouldn’t work in his situation, and why a ground-up solution like this is more appropriate for his farm. This post is largely an overview of his system, but some of his other posts go into more detail about things like integrating temperature sensors, rainfall monitoring, controlling irrigation systems, and plenty of other farm automation tasks that are useful for any farmer or gardener.
We’ve also seen some other projects of his here like this project which converts a common AC sprinkler system to an easier-to-use DC system, and a DIY weather station that operates in the 915 MHz band. He’s been a great resource for anyone looking to have technology help them out with their farm or garden, but if you’re just getting started on your green thumb be sure to take a look at this starter guide as well.
Nexo, a crypto lending platform, agreed to pay a $500,000 penalty after California regulators said it made thousands of loans without the proper state license.
According to the California Department of Financial Protection and Innovation, the actions involved loans backed by crypto assets and raised concerns about how the company evaluated borrowers.
The DFPI found that Nexo issued at least 5,456 consumer and commercial loans from July 2018 through November 2022 to residents in California.
Reports have disclosed that the company did not adequately check whether borrowers could repay the loans, leaving consumers exposed to risky lending. The agency called those practices unlawful under state consumer finance rules.
Nexo Must Move California Funds To Licensed Affiliate
As part of the remedy, Nexo will be required to transfer funds held for Californians to its US-based affiliate that holds a valid license, Nexo Financial LLC, within 150 days.
The move is meant to ensure customers’ money is under a properly regulated entity. The DFPI also required other compliance steps to prevent similar problems in the future.
This is not the first time Nexo has faced enforcement. Based on reports, the firm previously reached settlements that included roughly $45 million in penalties during actions taken in 2023.
Regulators around the country have been paying closer attention to crypto lending, and this decision signals they expect the same consumer protections that apply to traditional lenders to apply to platforms using digital assets.
Consumers who took loans secured with crypto may now see their accounts handled differently while the transfer takes place. Some borrowers might face changes in terms or servicing.
Industry observers say this kind of oversight can push companies to tighten underwriting and documentation. At the same time, some users worry that more rules could limit their access to certain crypto services.
Regulators Emphasize Borrower ProtectionsAccording to the DFPI, California law requires lenders to assess a borrower’s capacity to repay loans and to hold the right licenses before they are allowed to do business with state residents.
By labeling the conduct unlawful, the agency signaled that loan decisions driven primarily by crypto collateral do not exempt a lender from basic checks on repayment capacity. The penalty and the corrective measures aim to close gaps that might have allowed risky loans to go through.
A Cautious Road AheadThe $500,000 fine is modest compared with the scale of the broader crypto market, yet regulators say penalties are only one tool. They added that transfers to licensed entities and stronger internal controls are key to protecting consumers.
Featured image from unsplash, chart from TradingView
Crypto analyst TARA has predicted that the Bitcoin price will still rally despite bearish signals that have surfaced. She highlighted why the flagship crypto could reach this level and what could happen once it touches the price target.
In an X post, TARA opined that the Bitcoin price will reach $99,300, even though the flagship crypto is printing a bearish candlestick. She stated that BTC wants to touch this price target before it retraces deeper so that the correction does not break the critical support at $90,000. The analyst added that retracement levels for BTC will continue to be adjusted, with the new 2026 high above $97,000, while revealing subwaves on the way to the full target at $103,000.
Notably, crypto traders are currently betting on the Bitcoin price rallying past the $99,000 level and reaching the psychological $100,000 level. Polymarket data shows a 48% chance that BTC will rally to $100,000 this month. This follows the flagship crypto’s recent rally from around $92,000 to above $97,000 following the release of the soft CPI inflation data earlier this week.
The spot Bitcoin ETFs have also contributed to the Bitcoin price surge to start the year. In an X post, Bloomberg analyst Eric Balchunas highlighted that ETFs recorded net inflows of $843 million on January 14 and now boast 1-week net inflows of $1 billion and $1.5 billion year-to-date (YTD). With BTC rallying to $97,000 after trading sideways towards the end of last year, Balchunas opined that the buyers may have exhausted the sellers.
In his latest blog post, BitMEX co-founder Arthur Hayes predicted that the Bitcoin price could sustain this rally as dollar liquidity rapidly increases. Hayes expects dollar liquidity to increase as U.S. President Donald Trump finds more ways to inject liquidity into the economy. The BitMEX co-founder highlighted how Trump plans to lower mortgage rates, which could cause Americans to borrow more.
Hayes also mentioned that the liquidity in 2025 didn’t support crypto portfolios, which is why the Bitcoin price underperformed. He urged market participants not to draw wrong conclusions from the 2025 underperformance, as it was always a liquidity story rather than a cyclical bear market, as some analysts suggested.
More liquidity could also flow into the market as Trump nominates a rate-cut advocate to replace Fed Chair Jerome Powell. This could lead to larger rate cuts, which would be bullish for the Bitcoin price and the broader crypto market.
At the time of writing, the Bitcoin price is trading at around $95,300, down in the last 24 hours, according to data from CoinMarketCap.
If you’ve ever tried to build a serious financial product, screener, dashboard, or data pipeline, you already know the uncomfortable truth:
Getting financial data is easy.
Getting reliable fundamental data is not.
Most projects start the same way:
Then reality hits:
At that point, the problem is no longer technical.
It’s architectural.
That’s why choosing the right Fundamental Data API matters.
A Fundamental Data API provides structured, long-term financial information about assets, not just prices.
Unlike market data APIs (OHLC, ticks, volume), fundamental data answers deeper questions:
Derived data is what you build on top of fundamentals.
Examples:
Derived data is only as good as the raw fundamental data behind it.
If the base data is inconsistent, your models will be too.
These tools work for experiments, not for systems.
This is an architectural decision, not a feature checklist.
EODHD APIs scale from scripts to full platforms without changing your data model.
LoginGET https://eodhd.com/api/fundamentals/{SYMBOL}?api_token=YOUR_API_KEY&fmt=jsonSymbol examples:
import requests
import os
API_KEY = os.getenv("EODHD_TOKEN")
BASE_URL = "https://eodhd.com/api"
def get_fundamentals(symbol):
url = f"{BASE_URL}/fundamentals/{symbol}"
r = requests.get(url, params={
"api_token": API_KEY,
"fmt": "json"
})
r.raise_for_status()
return r.json()
stock = get_fundamentals("AAPL.US")
print(stock["General"]["Name"])
print(stock["Highlights"]["MarketCapitalization"])
print(stock["Valuation"]["TrailingPE"])Use cases
ETFs require look-through analysis, not just price tracking.
etf = get_fundamentals("SPY.US")
print(etf["General"]["Name"])
print(etf["ETF_Data"]["Holdings"].keys())Use cases
fund = get_fundamentals("SWPPX.US")
print(fund["General"]["Name"])Use cases
Indices are not just numbers.
Correct index analysis requires:
Using current constituents for past analysis introduces look-ahead bias.
This is essential for:
Crypto fundamentals are project-level, not just price-based.
btc = get_fundamentals("BTC-USD.CC")
print(btc["General"]["Name"])
print(btc["Statistics"]["MarketCapitalization"])
print(btc["Resources"]["Links"]["source_code"])Use cases
Typical integrations:
That’s a feature, not a drawback.
Economic and structural information about an asset, not its price.
Metrics or scores calculated from raw fundamental data.
Yes. That’s one of the main strengths of EODHD APIs.
Yes, when sourced from an official and stable API.
If you’re looking for a Fundamental Data API that lets you:
Then EODHD APIs provide a clean and professional foundation.
Access the EODHD Fundamental Data API with a discount:
Fundamental Data API: How to Extract Stock, ETF, Index, Mutual Fund, and Crypto Data (Step-by-Step… was originally published in Coinmonks on Medium, where people are continuing the conversation by highlighting and responding to this story.
