London-based Embedd has raised $2.7m (£2m) to develop its build capabilities of infrastructure for physical AI and expand its work with semiconductor partners.
Embedd’s platform creates a digital twin of chips and uses AI agents to generate the integration code engineers would otherwise write by hand. The startup says the approach lets manufacturers ship production-ready chip software up to six times faster.
Ukrainian founders Michael Lazarenko, Maxim Gorinov and Valentin Gololobov were running a hardware company when Covid-era shortages forced them to source whatever chips they could find and rewrite their software from scratch each time.
Each new chip meant weeks of manual integration work. Russia’s invasion of Ukraine disrupted their supply chain again – and this pattern convinced them the problem would only get worse as physical AI scales.
Seedcamp, the lead investor of the pre-seed round, says intelligent machines are built using chips from dozens of different manufacturers, and none of them are able to communicate with each other. Wiring them up means engineers writing pages of custom integration code for every single chip by hand – and this is the bottleneck Embedd is building to solve.
In addition to Seedcamp, the round also included participation from Cocoa, Connect Ventures, 2100 Ventures, Vesna Capital, U.ventures, Underline Ventures, Common Magic and Roosh Ventures.
“The next wave of AI will power factories, vehicles, robots and critical infrastructure, but today, every change in hardware creates huge complexity for software teams and that friction is already massively slowing innovation,” says Lazarenko.
“We built Embedd to address exactly that, and we’re thrilled to have the backing of Seedcamp as we scale.”
