Blacksmith Series C: $66M to Build Trust in AI Agent Code

Blacksmith, the continuous-integration cloud startup, announced on October 21 a $66 million Series C funding round led by Sequoia Capital that values the company at $1 billion, according to TechCrunch. The Blacksmith Series C brings the company's total funding to $125 million and arrives only two months after a $45 million Series B that had valued the startup at $550 million. A year ago, the company was worth $60 million after a $10 million Series A.

The reason for the breakneck fundraising pace is written into the announcement itself. With AI coding agents now producing a large and growing share of the world's code, the bottleneck in software has moved from writing code to validating it — and Blacksmith's dedicated CI cloud is positioned to be the infrastructure that does the checking, according to the company's release carried by PRNewswire.

The burden moved from writing to validating

For years, writing code was the slow part of shipping software. Every change still has to be built and tested before it can be merged, a step known as continuous integration, or CI. According to the company's release, most companies run CI on general-purpose cloud infrastructure that was never designed for the job. As developers push more code, teams hit longer waits for compute and slower builds on older-generation CPUs, with constant fights between over- and under-provisioning.

AI coding agents made the problem dramatically worse. Anthropic reported last month that its own CI volume grew 25 times in six months as Claude took over most of its code output, according to the release. Developers who used to push three or four pull requests a day are now pushing orders of magnitude more, and their CI infrastructure cannot keep up — a point Blacksmith co-founder and CEO Aditya "JP" Jayaprakash made in a statement accompanying the Blacksmith Series C.

That framing is the bet at the center of this round. "As agents write more of the world's software, the question that matters is whether it can be trusted," Jayaprakash argued in the announcement, as reported by TechCrunch. The comment connects directly to the trust question facing every team shipping agent-produced code — the same question that surfaced this week when OpenAI disclosed that its own AI agents had weakened safety monitors. Generating code is the easy part now. Proving it works is the business.

Why Sequoia is paying up now

Investors are chasing the momentum. Blacksmith says its annual recurring revenue has grown 17 times since January, according to PRNewswire's account of the announcement, and the platform has now run more than 90 million CI jobs, or individual automated code checks. More than 10,000 startups and enterprise companies run builds and tests on the platform, and its customer list includes the fintech Mercury and the meeting-notes company Granola.

The round drew participation from Premji Invest alongside existing investors Peak XV Partners, GV and Y Combinator, according to the release. That Sequoia led the Blacksmith Series C after Peak XV had led the Series B two months earlier signals how quickly the largest firms are racing to own the agent-build toolchain. The company plans to use the new capital to grow its team and expand its compute footprint.

Blacksmith's approach is deliberately physical. Rather than renting capacity from the big cloud providers, the company bought its own servers and built a dedicated CI cloud spanning multiple global locations, according to SiliconANGLE's reporting. It writes its own software to run the jobs, and one of its main advantages is automating much of the once-manual setup work required before each new piece of code can be tested.

Where validation sits in the agent stack

The raise lands at a moment when every layer of the AI-agent economy is racing to prove itself safe to use. Retailers and platforms are building the front doors for agents — Walmart opened its Agent Gateway to shopping agents this week — while labs and enterprises are grappling with what agents do once inside the gates.

Blacksmith is playing one layer down: the plumbing that decides whether machine-written code can merge into production. Jayaprakash's blog post announcing the round argued that before AI coding assistants, most developers spent about half their time writing code and the other half waiting for it to build and pass tests. Tools like Cursor, Claude and Codex made creating new code, in his framing, extremely easy — so the scarce resource became the test that says the code is safe.

That scarcity is why a company that simply builds and tests code is now worth a billion dollars. The Blacksmith Series C does not just fund servers and headcount; it funds the trust layer of an industry where software authorship is increasingly machine work. As agent output floods repositories, the question for the next funding cycle may not be who writes the code, but whose checks you trust to let it ship.