.jpeg)
Compute is the largest market of this AI cycle, and is fragmenting rapidly. As AI adoption accelerates, we expect to see highly specialized, high-performance compute infrastructure and software platforms designed to excel at specific workloads. Blacksmith is building such a platform for code builds, validation, and integration as AI-generated code explodes. Thousands of AI-native startups and digital enterprises already rely on Blacksmith, with many more discovering its capabilities every week. We are thrilled to lead Blacksmith's $45M Series B alongside Google Ventures, Y Combinator, and a group of exceptional angel investors.
Code generation represents the most compelling opportunity of this AI cycle. AI-native developers are leveraging specialized harnesses like Opencode and Cmux to run long-horizon, simultaneous coding agents across multiple clouds and producing orders of magnitude more code than ever before. However, this shift creates new downstream bottlenecks: code slop, poor integration, and inefficient execution at runtime. The core challenge in software development has moved from generating code to making that code performant, reliable, and production-ready.
Enter Blacksmith. When we first met JP last September, our conviction in the AI code generation space was already high. But we were struck by JP’s clarity of vision and the intentional, nuanced architecture choices Blacksmith had made to solve this exploding bottleneck.
JP walked us through how Blacksmith built a specialized software stack on top of proprietary, bare-metal infrastructure to deliver high-performance compute specifically engineered for code validation and integration. By tailoring the hardware-software stack to these exact workloads, they have made CI/CD orders of magnitude faster and significantly cheaper. But what excited us most was the broader vision: building the definitive high-performance compute platform specifically designed for coding-related downstream applications.
We were super intrigued. JP and his cofounders are insanely ambitious, and that resonates with us. The company was still small at the time, with a single-digit team size (our favorite stage to partner) and single-digit million ARR, but we noticed how AI-native startups were rapidly adopting the platform. We love young companies with product-led growth that are still underinvested in sales. After speaking with customers and hearing customer feedback, it was an easy decision to partner with Blacksmith. In fact, as we spent more time with the team, our conviction only grew stronger. We upsized our investment after signing the termsheet to invest even more before closing.
We are thrilled to have the opportunity to partner with JP, Aayush and Maru and believe they will build a foundational infrastructure company in the autonomous coding era.
