{"slug": "embedd-s-ukrainian-founders-raise-2-7m-to-automate-the-chip-integration-grind", "title": "Embedd's Ukrainian founders raise $2.7M to automate the chip integration grind", "summary": "Embedd, a London-based developer-tools company founded by Ukrainian entrepreneurs Michael Lazarenko, Maxim Gorinov, and Valentin Gololobov, has raised $2.7 million (€2.31 million) in a pre-seed round led by Seedcamp to automate chip integration by converting semiconductor documentation into production code. The funding, reported by The Recursive on August 24, follows the founders' previous hardware venture Agrolabs, which was disrupted by chip shortages and Russia's invasion, leading them to develop software that generates drivers and configuration files for embedded systems without relying on AI for code generation.", "body_md": "# Embedd's Ukrainian founders raise $2.7M to automate the chip integration grind\n\n**Seedcamp led the pre-seed round as Embedd expands software that turns semiconductor documentation into production code.**\n\nBy [RuntimeWire Staff](/author/runtimewire-staff)\n· Published\n\nPrimary source: [The Recursive](https://www.therecursive.com/embedd-raises-eur2-31m-to-build-software-infrastructure-for-physical-ai/)\n\n## Why it matters\n\nPhysical AI depends on unglamorous integration work between chips and software. Embedd is betting that automating it can give chipmakers faster distribution into robotics, vehicles and industrial systems.\n\n[Michael Lazarenko](https://www.eenewseurope.com/en/ai-for-hardware-software-integration-embeddit/?ref=runtimewire), Maxim Gorinov and Valentin Gololobov have raised $2.7 million for [Embedd](https://embedd.it/company?ref=runtimewire), the London developer-tools maker they founded after semiconductor shortages and Russia's invasion dismantled their previous hardware operation. [The Recursive reported on August 24th](https://www.therecursive.com/embedd-raises-eur2-31m-to-build-software-infrastructure-for-physical-ai/?ref=runtimewire) that Seedcamp led the 2.31 million euro pre-seed round.\n\nThe three Ukrainian founders arrived at Embedd through the least theoretical version of its problem. Lazarenko and Gorinov previously built Agrolabs, a Dnipro-based maker of wireless sensors and control systems for greenhouses. [AIN reported in 2020](https://en.ain.ua/2020/01/31/agrolabs-case/?ref=runtimewire) that Agrolabs had reached $345,000 in sales and pre-orders, entered the Brinc accelerator in Hong Kong and taken $100,000 from Artesian Venture Partners.\n\nThen hardware stopped cooperating. Chip shortages forced Agrolabs to replace components and rewrite firmware around the substitutes. Russia's invasion later took the production facility, Lazarenko [told EE News Europe](https://www.eenewseurope.com/en/ai-for-hardware-software-integration-embeddit/?ref=runtimewire). That sequence gave the founders Embedd's thesis: much of chip integration is a documentation and data problem disguised as repetitive software engineering.\n\n### Turning datasheets into software\n\nEmbedd ingests datasheets, technical reference manuals and related files, then converts their contents into structured models of a component's registers, peripherals and software-facing behavior. Engineers can use those models to generate drivers, hardware abstraction layers, board-support packages and device-tree configurations for embedded operating systems.\n\nThe distinction between extraction and generation matters. Lazarenko has said Embedd uses AI to retrieve and organize information from inconsistent hardware documents, while relying on deterministic systems to produce code. Embedd's [technical documentation](https://docs.embedd.it/pages/drivertargets?ref=runtimewire) says its driver-generation process does not use AI, an important constraint in markets where a plausible-looking error can break a physical device or undermine safety certification.\n\nEmbedd also offers a [vendor-neutral MCU and MPU configurator](https://embedd.it/company?ref=runtimewire) that lets developers assign pins and peripherals, identify conflicts and generate device trees for Zephyr inside a browser or VS Code. Chip vendors usually provide their own configuration environments and formats, leaving engineering groups to relearn tooling whenever they change suppliers. Embedd is betting that a common representation across vendors can make that switching work less painful.\n\nThat product direction grew out of customer feedback. Embedd initially concentrated on peripheral devices, Lazarenko told EE News Europe, before users pushed it toward microcontrollers and support for operating systems including Zephyr, Linux and AUTOSAR. The founders are effectively moving Embedd upstream, from automating individual drivers to assembling a larger share of the software package needed to make a board usable.\n\n### Seedcamp backs the layer between chips and applications\n\n[Seedcamp](https://seedcamp.com/?ref=runtimewire) led the round, with participation from [Cocoa](https://www.cocoa.vc/?ref=runtimewire), [Connect Ventures](https://www.connectventures.co/companies/embedd?ref=runtimewire), [2100 Ventures](https://www.2100.vc/?ref=runtimewire), [Vesna Capital](https://vesna.capital/eng/?ref=runtimewire), [u.ventures](https://u.ventures/about-us/?ref=runtimewire), [Underline Ventures](https://underline.vc/?ref=runtimewire), [Common Magic](https://www.commonmagic.xyz/?ref=runtimewire) and [Roosh Ventures](https://roosh.vc/about/?ref=runtimewire). Embedd plans to use the capital to develop the platform and work with additional semiconductor manufacturers.\n\nThe investor list also shows that Embedd's financing history predates this week's announcement. Connect Ventures lists 2023 as its investment year, the same year that UK records show [Embedd Limited was incorporated](https://find-and-update.company-information.service.gov.uk/company/14618092?ref=runtimewire). The new round puts a larger number around a founder-investor relationship that began while Embedd was still taking shape.\n\n[Seedcamp General Partner Carlos Eduardo Espinal](https://www.therecursive.com/embedd-raises-eur2-31m-to-build-software-infrastructure-for-physical-ai/?ref=runtimewire) framed the investment around the software required for AI systems operating in factories, vehicles and other physical settings. That thesis gives Embedd room to sell picks and shovels into the physical AI funding cycle without building robots, vehicles or chips itself. Embedd instead has to persuade semiconductor manufacturers that better software support can influence whether engineers select and keep using their components.\n\n### Embedd puts Microchip on its customer list\n\nEmbedd says it is working with multiple semiconductor manufacturers and identifies Microchip Technology as its latest example, with Embedd enabling support for the Zephyr real-time operating system. [Microchip's developer materials cover Zephyr RTOS-based embedded solutions](https://www.microchip.com/en-us/tools-resources/develop/zephyr?ref=runtimewire).\n\nEmbedd also joined the [Zephyr Project as a Silver member in February](https://zephyrproject.org/zephyr-project-welcomes-new-members-february-2026/?ref=runtimewire). In that announcement, Lazarenko said embedded development needed greater standardization around hardware abstraction and described Embedd's goal as making hardware integration repeatable.\n\nZephyr is a useful proving ground because its hardware coverage depends on a long tail of board definitions, drivers, bindings, tests and documentation. Adding nominal support for a chip is only the beginning. Production users need the relevant peripherals to work across a device family and remain maintained as the operating system changes. That creates a recurring workload for chipmakers and a possible recurring business for Embedd.\n\nEmbedd says customers have delivered production-ready chip software [up to six times faster](https://www.therecursive.com/embedd-raises-eur2-31m-to-build-software-infrastructure-for-physical-ai/?ref=runtimewire) using its platform. That is a company-supplied ceiling rather than a disclosed benchmark, and Embedd has not attached a test method or representative project to the figure. The stronger evidence is the direction of the product: Embedd has moved from a founder pain point to an enterprise toolchain, joined the Zephyr organization and secured a named engagement around Microchip's software support.\n\nThe round gives Lazarenko, Gorinov and Gololobov capital to test whether that work becomes infrastructure rather than another engineering service hidden behind an AI label. Their advantage is unusually concrete. They built and manufactured connected hardware, endured component substitutions and rewrote firmware when the supply chain changed underneath them. Embedd is the software layer they needed at Agrolabs and could not buy.", "url": "https://wpnews.pro/news/embedd-s-ukrainian-founders-raise-2-7m-to-automate-the-chip-integration-grind", "canonical_source": "https://runtimewire.com/article/embedd-ukrainian-founders-raise-2-7m-chip-integration", "published_at": "2026-08-25 05:50:22+00:00", "updated_at": "2026-08-25 06:13:24.287044+00:00", "lang": "en", "topics": ["developer-tools", "artificial-intelligence", "ai-tools"], "entities": ["Embedd", "Seedcamp", "Michael Lazarenko", "Maxim Gorinov", "Valentin Gololobov", "Agrolabs", "The Recursive", "EE News Europe"], "alternates": {"html": "https://wpnews.pro/news/embedd-s-ukrainian-founders-raise-2-7m-to-automate-the-chip-integration-grind", "markdown": "https://wpnews.pro/news/embedd-s-ukrainian-founders-raise-2-7m-to-automate-the-chip-integration-grind.md", "text": "https://wpnews.pro/news/embedd-s-ukrainian-founders-raise-2-7m-to-automate-the-chip-integration-grind.txt", "jsonld": "https://wpnews.pro/news/embedd-s-ukrainian-founders-raise-2-7m-to-automate-the-chip-integration-grind.jsonld"}}