CodeRabbit secured $143 million to analyze the code generated by AI.

      Code review is the process in which someone examines a change before it is deployed. A developer suggests an edit, opens a pull request, and a coworker reviews it for errors and security vulnerabilities. This task is often slow and unglamorous. Until recently, the process was manageable because typically one person wrote each change and another reviewed it. However, this balance has changed. Coding agents now work for extended periods and submit pull requests with minimal human input.

      On Wednesday, CodeRabbit announced it had secured $143 million to address this issue. Bloomberg reported the funding round prior to the announcement. Atomico and Smash Capital co-led the investment. Chief executive Harjot Gill told Axios Pro that the post-money valuation stands at $1.5 billion. He and Gur Singh established the company in San Francisco in 2023. Gill previously sold a startup, Netsil, to Nutanix in 2018.

      The lack of human oversight on the code produced by agents is a concern. The company attributes the rising issue to external factors. According to the chief operating officer, GitHub is expected to document 14 times the number of commits this year. In companies that are early adopters of coding agents, these agents are responsible for 35% of pull requests, Jellyfish reports.

      The demand for coding agents exceeds the supply. Rather than allow the usage to spiral, GitHub Copilot paused new registrations. "AI has made code abundant," CodeRabbit stated on X. "Trust has now become the limiting factor."

      The product automatically reviews changes, identifying bugs, security risks, and maintenance issues, while also explaining what each modification entails. The company claims it conducts over two million reviews weekly for more than 17,000 clients, including Nvidia, BMW, Adyen, Indeed, JFrog, and Trivago.

      A European fund co-led this American funding round. Atomico, based in London, rarely co-leads U.S. growth rounds of this magnitude. Partner Luca Eisenstecken will join the board. He stated, "As AI becomes a crucial part of global infrastructure, organizations will increasingly require independent governance layers to validate software, irrespective of the model used in its creation."

      The investment is driven by customer demand. CodeRabbit has established an office in Moorgate, London, and currently employs 50 people across the UK and EU, with plans to double that workforce in Europe before expanding into Japan and Singapore. BMW i Ventures also participated in the funding round. Managing partner Kasper Sage highlighted that car manufacturers need dependable software and efficient engineering teams. BMW is already a CodeRabbit client, so the fund is supporting a product used by its own engineers.

      Several investors have roles on both sides of the equation. Datadog is a new investor, while Nvidia was already involved and is also a customer. This connection is both technical and financial. CodeRabbit post-trained Nvidia's Nemotron 3.5 Lightning model for one of its busiest routing tasks. Existing backers CRV, Scale Venture Partners, Flex Capital, and Pelion Venture Partners also renewed their support.

      CodeRabbit announced a new product category along with the funding, termed Agentic Change Management. The founders' blog was more blunt, declaring, "Issue tracking is dead." Their argument is that tools like Jira were designed during an era when coding was costly, prompting teams to plan first and code second. With AI, the sequence has flipped; code often exists before its value has been determined.

      Anyone can initiate a change now, as OpenAI has promoted Codex to non-developers for similar reasons. Three new tools have emerged as a result: Triage evaluates incoming pull requests based on value, urgency, and risk, automating those deemed low-risk; Change Stack organizes a substantial agent-generated diff into manageable layers for human review; and CodeRabbit Security scans already deployed code.

      The emphasis on security was a key factor in the funding round. Large language models have reduced the cost of identifying vulnerabilities, benefiting both defenders and attackers. This is why the security product was introduced alongside the funding instead of afterward. Brad Twohig, founding partner at Smash Capital, asserted, "Security now requires a continuous system capable of reasoning across the repository, verifying whether a threat is genuine, and integrating fixes back into the development workflow."

      The supply chain reflects similar challenges. An npm worm impacted developer packages this year.

      However, the market is smaller compared to the company. The global code review services market is estimated at around $3.04 billion this year, with predictions of reaching $4.79 billion by 2032, according to figures from 360iResearch cited by Tech Funding News. Meanwhile, CodeRabbit is valued at $1.5 billion, which is about half of the entire market it currently addresses.

      This valuation hinges on the assumption that code review will evolve from a developer tool into a core infrastructure component. Gill asserts that the demand supports this shift. "It's almost like a land grab