The AI coding boom created a strange new problem: machines now write code faster than anyone can check it. Blacksmith, a two-year-old startup that runs the build-and-test infrastructure standing between AI-written code and your production servers, announced a $66 million Series C led by Sequoia Capital on October 21 at a $1 billion valuation – nearly double its $550 million mark from about two months ago.
Key takeaways
- Blacksmith raised a $66 million Series C led by Sequoia at a $1 billion valuation, with Premji Invest, Peak XV, GV and Y Combinator participating; total funding is about $125 million.
- The company runs continuous integration – the automated builds and tests every code change must pass – on its own hardware, and says AI coding agents have exploded that workload.
- Company-reported traction: CI jobs grew from 5 million to 90 million per week since January, ARR grew 17x, and more than 10,000 companies use the platform.
- The valuation roughly doubled in about two months, after a $45 million Series B at $550 million in August, per Fortune.
The round and the velocity
The Series C brings Blacksmith’s total funding to roughly $124.5 million, per Fortune’s exclusive. The pace is the story: a $550 million valuation in August, $1 billion in late October. Sequoia partner Isaiah Boone said the firm backed Blacksmith because it “understood early that as AI agents wrote more code, the infrastructure to build and test it would need to change,” per the company’s announcement.
The traction claims behind the repricing, all company-reported: weekly CI jobs up from 5 million in January to more than 90 million; annual recurring revenue up 17x this year and now in the tens of millions, per CEO Aditya Jayaprakash; customers up from about 5,000 in August to more than 10,000, including LangChain, Supabase, Granola, Handshake and Ashby; the largest customers spend more than $1 million a year.
Why validation became the bottleneck
Continuous integration is the least glamorous layer in software: every proposed change gets built and tested before it merges. For years it was background noise, because humans could only write so much code. Then AI coding agents removed that limit. Jayaprakash’s summary: developers who pushed three or four pull requests a day now push orders of magnitude more, and the CI infrastructure cannot keep up. Anthropic reported last month that its own CI volume grew 25x in six months as Claude took over most of its code output – a data point Blacksmith’s announcement leans on heavily.
Blacksmith’s answer is unglamorous too: own the machines. Instead of renting general-purpose cloud servers, it buys its own hardware – tens of thousands of machines, per Jayaprakash – installs them in data centers across the US and Europe, and runs custom software tuned for the spiky CI workload, where one customer’s demand can jump from zero to tens of thousands of CPUs and back. The company claims jobs complete twice as fast at 60% lower cost on average; those are its numbers, not benchmarks we verified.
The founding team is a systems pedigree bet: Jayaprakash, Aayush Shah and Aditya Maru met as University of Waterloo undergrads, with Shah and Maru early engineers at Cockroach Labs and Jayaprakash at Faire. They founded Blacksmith in 2024 on the thesis that testing code, not writing it, would become the constraint. That thesis aged well.
What it means if you run a business
You do not need to buy CI infrastructure to learn from this round. First, every gold rush reprice its tools: as AI made code cheap, verification got expensive, and the money followed the new constraint. Whatever AI makes abundant in your business, something downstream just became scarce – that scarcity is where the pricing power moved. Second, watch the “own the bottleneck” model: Blacksmith bought hardware against a cloud-default industry because its workload is uniquely spiky. The general-purpose default is not always the answer once you understand your own demand curve. Third, expect your vendors’ costs to show up here: AI-written code still needs testing, review and accountability, and someone pays for that compute.
Frequently asked questions
What does Blacksmith do?
Blacksmith operates continuous integration infrastructure – the automated builds and tests every code change must pass before merging – on its own hardware in data centers across the US and Europe. Customers switch with a one-line change to their workflow file, per the company.
How much has Blacksmith raised?
Blacksmith announced a $66 million Series C led by Sequoia Capital at a $1 billion valuation on October 21, 2026, about two months after a $45 million Series B at $550 million. Total funding is roughly $125 million.
Why is code validation suddenly a big market?
AI coding agents let developers produce far more code than before, and every change still has to be built and tested. Anthropic reported its own testing workload grew 25x in six months as Claude took over most of its code output, making validation the new bottleneck.
Who founded Blacksmith?
Aditya “JP” Jayaprakash, Aayush Shah and Aditya Maru, who met as undergraduates at the University of Waterloo. Shah and Maru were early engineers at Cockroach Labs; Jayaprakash was an engineer at Faire. The company has about 40 employees and plans to reach roughly 75 by year-end, per Fortune.
Method: Blacksmith’s October 21 announcement via PRNewswire and Fortune’s same-day exclusive, both fully fetched October 21. All customer counts, ARR, growth, speed and cost figures are company-reported and labeled as such. This article is not investment advice. Photos are illustrative data-center scenes, not Blacksmith facilities.



