Seedcamp has led a $2.7 million pre-seed into Embedd, a London company built on the argument that the bottleneck in shipping robots, vehicles and medical devices is no longer the silicon but the software written to talk to it.

The round was announced on Aug 24, 2026 and reported as $2.7 million, or EUR 2.3 million; no GBP figure was published, despite the company being UK-based. Seedcamp led, with participation from Cocoa, Connect Ventures, 2100 Ventures, Vesna Capital, U.ventures, Underline Ventures, Common Magic and Roosh Ventures. Embedd was founded by Michael Lazarenko, who serves as chief executive, alongside Maxim Gorinov and Valentin Gololobov. No valuation was disclosed, no founding year was given, and none of the coverage records earlier funding. A global deal feed lists the round at $3 million.

What the company does

Embedd platform converts chip documentation into structured, machine-readable models, then generates configuration and code from them. In practice that means parsing datasheets, errata and SVD files, handling microcontroller configuration and device-tree generation, and emitting code with accompanying documentation. Output targets bare-metal, Zephyr, Linux and AUTOSAR, and the tooling plugs into a VS Code extension and existing CI/CD pipelines.

The company frames this as infrastructure for physical AI, using digital twins of hardware and AI agents to remove the manual firmware work that sits between a chip and a working product. Its stated focus is on sectors with fixed architectures, thin existing support and heavy compliance burdens — automotive, aerospace, medical, defence and consumer.

The numbers behind the round

Embedd says customers can deliver production-ready software for a chip up to six times faster than by hand, compressing integration work that typically runs four to six months into roughly three weeks. The platform launched commercially in April 2026, and the company has signed contracts with several semiconductor firms, including Microchip Technology, where the work covers Zephyr support.

The money is earmarked for platform development, deeper partnerships with chipmakers, and scaling to meet demand. For a pre-seed, the customer list is the more telling number: semiconductor vendors are the buyers, not the end device makers.

Why the timing matters

The commercial logic is set out bluntly by Microchip vice-president Rodger Richey, who said the competitive question in embedded is no longer whose silicon is fastest, but whose silicon is easiest to build on. That reframing is what makes a developer-tooling company interesting to chip vendors rather than merely useful to their customers.

Lazarenko makes the demand-side case: the next wave of AI will power factories, vehicles, robots and critical infrastructure, but every hardware change creates fresh complexity for software teams. Seedcamp Carlos Eduardo Espinal described the problem as fundamental across robotics, manufacturing, healthcare and automotive. The founders are Ukrainian, and their previous hardware company was disrupted by Russia invasion — context that explains a team rebuilding in London around software rather than devices.

Source

EU-Startups — "London-based Embedd raises EUR 2.3 million to build the software infrastructure for physical AI"