CopperPilot won Accel’s Anthropic Dev Day, an event focused on teams building serious AI products—not slide-deck chatbots, but software that changes how work gets done.
The competition drew hundreds of participants. Winning validated something we felt in customer conversations long before demo day: hardware engineers are ready for a copilot, but only if it respects how unforgiving physical design is.
Our pitch was straightforward. LLMs are good at language. Hardware design needs language plus structure—net names, pin maps, BOM constraints, firmware interfaces, thermal limits, and review checklists. CopperPilot connects those structures so the model is answering questions about your board, not a generic training snapshot.
Judges cared about three things: depth of workflow integration, quality of outputs on real examples, and whether the product could become daily infrastructure for a team. That maps cleanly to how we prioritize the roadmap.
Winning does not change the mission—it raises the bar. More teams will look at AI for hardware in the next year. The winners will be products that earn trust on real projects, not prototypes that look good in a two-minute video.
Thank you to Accel, Anthropic, and the other founders who pushed us to sharpen the story. We are back to building.