Our Investment in Code Metal: AI Development for Hardware, Defense, and Regulated Industries

AI is helping teams move from idea to prototype at record speed. But getting these prototypes to work reliably and consistently on hardware remains a challenge. In sectors with tight regulations and complex edge computing systems, modern developer tools often fall short. That disconnect hits hardest in places where error is high stakes: defense, automotive, and anywhere else compliance isn't optional.

Verifiable AI code translation for edge environments

Enter Code Metal: provably correct AI code translation for when both speed and accuracy are mission-critical. CEO and co-founder Peter Morales first came up with the idea while working at BAE Systems, where he was tasked with optimizing machine learning algorithms to help the F-35 jet fighter counter enemy radar.

While engineers had already written algorithms to identify radar signals, getting them working in the field was another story. Code written in everyday programming languages must be translated into machine-level instructions to run on specialized chips in aircraft and other hardware. In environments with strict constraints like defense, developers often handle parts of the workflow manually to eke out every possible performance gain. Doing this firsthand sparked Peter's vision for how automation could accelerate workflows and save massive amounts of resources.

At its most basic, Code Metal is a translation platform for programmers: It enables developers to write code in a standard programming language like Python and then translates it to run on a given device's hardware. Testing is automated, ensuring that every line of code is robust, optimized, and compliant. Code Metal uses a mixture of LLM and traditional code processing methods to ensure each line of code is highly efficient-essential for devices that may have memory and processing constraints.

Code Metal's edge computing expertise

Code Metal already works with clients in the defense, automotive, and semiconductor industries, including container feeder operator X-Press Feeders and defense contractor L3Harris. They're on track to deliver eight figures in revenue this year and are helping their clients shrink product deployment times from months to days.

We couldn't think of a better team to tackle this challenge. Peter has a decade of experience as an AI research scientist across various labs, including MIT Lincoln Laboratory, where he spearheaded automation and robotics projects for the Air Force. His co-founder, Alex Showalter-Bucher, spent a decade at MIT Lincoln Laboratory and has worked with the Navy, Army, and Department of Homeland Security, among others.

In leading their Series A, we're continuing our tradition of partnering with teams who apply deep technical expertise to solve complex, high-impact problems. With Code Metal, we see the potential for advances that shape some of the world's most important industries and institutions for decades to come.