Circuit Mind thinks hardware design missed the productivity wave software rode. The London-based startup, which automates printed circuit board work, closed $25M and launched an Agentic Verification Engine that sends AI agents through schematics looking for faults. Electronics design, chief executive Tomide Adesanmi says, has not changed in 40 years.
The pitch rests on a limitation. General-purpose language models have never been trained on professionally designed circuits, so Circuit Mind wraps its agents around a patented physics-based engine that checks every design choice against electrical limits. The agents read the company’s machine-readable component models and a customer’s own rules, down to approved manufacturers.
The verification engine traces each connection and bus against pin tables and operating voltages, then critiques its own findings before an engineer sees anything.
Customer results are concrete. Los Alamos National Laboratory finished a medium-difficulty board in just over four hours, against an estimated 60 to 80 hours of manual effort, per a publicly released report. Schneider Electric saw more than 90 percent less effort exploring electronic architectures, according to its energy management chief technology officer Daniel Gheno. Users have cut bill-of-materials costs by 25 percent.
Norwest Venture Partners led an $18.5M Series A, with Lightscape joining as a new investor and Third Prime, Differential Ventures and Blackhorn returning. The remaining cash comes from a previously unannounced seed round. Total backing now reaches $33.4M.
