Embedder’s cover photo
Embedder

Embedder

Software Development

AI Agents for Firmware

About us

AI Agents for Firmware

Website
https://www.embedder.com
Industry
Software Development
Company size
2-10 employees
Headquarters
San Francisco
Type
Privately Held
Founded
2025

Employees at Embedder

View 13 employees at Embedder

or

By clicking Continue to join or sign in, you agree to LinkedIn’s User Agreement, Privacy Policy, and Cookie Policy.

See all employees

Locations

Updates

  • Embedder reposted this

    When I talk to developers and dreamers, they all point to hardware-in-the-loop testing as the holy grail of firmware development. But HIL testing has limits. There are only so many boards, only so many engineers, and only so many tests you can think to run. When you design tests to debug and optimize a device, you're usually working from three things: datasheets, field logs, and engineering intuition. Embedder's next release will change what you can do with all three. 1) Hardware-in-the-loop testing. Embedder runs QA and optimization tasks directly on your bench, driving every instrument against a single run, so each test is captured from every angle. And it doesn't just run them: using native context and our proprietary SDK, Embedder hypothesizes, tests, and iterates on your boards, correlating what each instrument sees to pinpoint and fix issues. 2) Field data. Whether you run your own crash reporting or pull it in through an API, Embedder folds real-world crash data into its context, so tests reflect how your devices actually fail in the field. 3) Multi-agent architecture. To break the one-developer-one-board bottleneck, we've built multi-agent testing. On a single board, the agents iterate autonomously — running experiment after experiment and refining firmware between runs. Add more boards and they scale out in parallel, one agent per board. Either way, they work while you sleep. When devices fall short or degrade in the field, everyone in the business feels it. These features are built to help you ship products faster, and better.

    • No alternative text description for this image
    • No alternative text description for this image
  • Embedder reposted this

    This past month the Embedder team visited Taipei, Hong Kong, London, Oslo, Stockholm, and Munich, meeting customers, partners, and investors face to face. This was my longest international business trip (30+ days), and I basically lapped the entire globe. I have a few reflections. 1) The appetite for AI is higher than ever, but the texture differs by region. In Asia the energy is around semiconductors and hardware. In Europe it's automotive and industrial. Everywhere, teams have stopped asking whether AI belongs in firmware and embedded development. They're asking how fast they can get it there. But there are also unique constraints: functional safety, confidentiality/IP protection, compliance, maintainability. Things that don't concern generic software buyers are deal-breakers in embedded systems. 2) I'm convinced that in-person is the only way to do big business as a startup, at least when you want to build solutions for industrials, suppliers, and OEMs. There's no replacement for getting on a plane and sitting across the table from someone. Embedder prefers live demos, because that's the reality of engineering. It's one thing to watch a live demo on a call, but nothing replicates the moment when a piece of metal comes to life on your desk. 3) Driving adoption of new tech inside a large enterprise isn't a solo effort. It's collaborative, and it almost always needs an innovative company in the room pushing things forward. What surprised me most: our strongest champions were often the most senior people. They've lived through enough platform shifts to know this is the next one, and they want to be first to it, not last. The next year of AI in embedded development is going to move fast. The teams that win will be the ones rethinking how they build, not just what they build with.

    • No alternative text description for this image
    • No alternative text description for this image
    • No alternative text description for this image
    • No alternative text description for this image
  • "Trigger on the heartbeat on LED 1, max sample rate." With one natural language prompt, Embedder connected to the oscilloscope over LAN, configured the trigger and timebase, ran the capture, and returned the waveform. A clear rising edge at 2 GSa/s. See how in under 2mins. Full walkthrough in comments.

  • Embedder reposted this

    Embedder was featured in Electronic Product Design & Test [EPDT] this week. I had a fantastic conversation with Mike Green, who asked something I've been thinking about for a while: does being young hurt us when selling into enterprise? It used to. A few years ago, "21-year-old with no industry experience" was a dealbreaker for most enterprise buyers in hardware. That's not really true anymore, and I think AI is the reason. With the right tools and a willingness to learn fast, you can now compress years of learning into weeks. The conversations we're in today look very different from the ones we were having 6 months ago.

    • No alternative text description for this image
  • Embedder reposted this

    Want to just say "Computer, optimize firmware energy consumption?" You can almost do that today! Nice work Embedder, and really cool to see Joulescope integration.

  • Embedder reposted this

    Start-Up Profile: Embedder Company name: Embedder Website: www.embedder.com Founded: ~12 months ago Location: San Francisco, USA Industry: Embedded systems Technology: AI-powered code generation EPDT initially got to learn about Embedder during the Embedded World conference/exhibition in Nuremberg earlier this spring. The emerging tech firm was co-founded by Ethan Gibbs (who now serves as its CEO) and Bob Wei (who holds the post of CTO) towards the end of their academic studies (when they were both still only a mere 21 years of age). Through the use of AI agents, its proprietary platform is able to automate the firmware code creation process - drawing upon information from disparate items of bulky and normally hard to consume documentation, with hardware-in-the-loop (HiL) based testing then being employed in order to validate the code that has been generated, before any subsequent iterations are undertaken. The upshot of all this is that development work can be squeezed into much shorter timeframes, with far less engineering resources needing to be allocated. Over 400 different microcontroller unit (MCU) models are already supported by the platform. The company has a diverse and steadily growing client base - with customers across North America, as well as several in Europe and Asia. It has already managed to pick up a couple of notable industry accolades too. Furthermore, it has been successful in raising several million dollars’ worth of venture capital money so far, with plans to commence its Series A funding round in either late-2026 or early-2027. 12 people are now working there (including 4 interns), with the expectation of reaching somewhere between 15 and... READ THE FULL PROFILE https://lnkd.in/eq-yKtNk Massimo Sonnino Sorisio Jonah Perry Samuel Siviero Aidas P. Bao Duong Reet Oberoi Alexander Sereno Jared Friedman Christ Xu Leo Gao Catia Lai Guy Forster Etelka Zdechovan Sandro Mark Aby Parsons Y Combinator

    • No alternative text description for this image

Similar pages

Funding

Embedder 1 total round

Last Round

Pre seed

US$ 500.0K

Investors

Y Combinator
See more info on crunchbase