CodeRabbit Raises $143M at $1.5B Valuation, Extending AI Code Review Into Agentic Change Management

The real bottleneck vibe coding created isn’t how fast code gets written — it’s how fast anyone can decide whether to trust it. Developers, product managers, marketers, background agents, issue trackers and production alerts can all now produce code and open a pull request, the request to merge a change into the real codebase and the gate where a colleague inspects it. The queue at that gate has grown past what teams can absorb.

Issue trackers like Jira assumed the opposite sequence. Engineering capacity was scarce and expensive, so work was curated and prioritized before anyone wrote a line. That order has flipped: code now exists before a team has agreed the work should happen at all. The backlog has moved from tickets to proposed code, which pushes triage, planning and governance downstream. The pull request becomes the auditable decision point — where a team judges whether a change clears the quality bar, how much risk it carries, whether it deserves human attention, and what it means for the wider system.

CodeRabbit announced on August 12 that it raised a $143 million Series C at a $1.5 billion valuation and introduced Agentic Change Management, a control layer for governing, understanding and shipping software created by people and agents alike.

From review to control layer

CodeRabbit validates changes using repository-wide context, organizational standards, pre-merge checks, accumulated team knowledge, and evidence from isolated test environments where the code is actually run. Coding agents work through its findings in automated fix-and-re-review loops before a human sees anything, so reviewers spend their attention on meaning and judgment rather than defect-hunting.

Three new capabilities extend that layer beyond validation. Triage scores incoming pull requests on value, urgency, risk, dependencies, readiness and reviewer fit, then routes consequential work to people, low-risk changes into automated workflows, and duplicate or unready work out of the queue. Change Stack drops the alphabetical file listing in favor of a guided walk through the contracts, domain behavior, integrations, tests and migrations a change touches, with blast-radius and architecture analysis showing how far the change reaches and where deeper scrutiny is required.

CodeRabbit Security extends the layer past merge. Full-repository scans and continuous monitoring hunt for vulnerabilities in code already running in production, adjudicate each finding against code evidence, and send proposed fixes back through the pull request process — so the same control layer that screens incoming changes keeps protecting the code after it ships. It runs on a proprietary codegraph, fine-tuned over years of code review, that dissects a codebase into trust boundaries and reasons across the whole repository. That targets vulnerability classes fixed rules and file-level pattern matching cannot express: authorization bypasses, insecure direct object references, business-logic flaws, and threats emerging from AI-enabled systems. The timing is deliberate — large language models have lowered both the cost and the expertise needed to find vulnerabilities, letting attackers and defenders alike sweep far more code and far more attack paths.

Co-founder and CEO Harjot Gill researched networking, streaming data analytics and distributed systems as a PhD candidate at the University of Pennsylvania, then turned that work into Netsil, a cloud observability startup acquired by Nutanix in 2018, where he served as senior director of technology. He founded FluxNinja before starting CodeRabbit. Co-founder Gur Singh led engineering and product teams building healthcare SaaS at Alegeus and elsewhere.

“Code changes now originate across the software organization, including from developers, non-technical personnel, as well as from coding agents to issue trackers, support systems, and production alerts. Every change creates a decision for the team,” Gill said. “We pioneered AI code review to give teams an independent system that validates the work before it ships. Agentic Change Management expands that layer to determine what deserves attention, explain each change’s impact, and continue monitoring the codebase after it ships.”

2.7x in eleven months

Atomico and Smash Capital co-led the round. New investors include BMW i Ventures, Datadog, Hirtle Callaghan, SineWave Ventures and Scenic Management, alongside existing backers CRV, Scale Venture Partners, Flex Capital, Pelion Venture Partners, Harmony Partners and Engineering Capital. Atomico partner Luca Eisenstecken joins the board.

The pace stands out. CodeRabbit raised a $16 million Series A led by CRV in August 2024 and a $60 million Series B led by Scale Venture Partners in September 2025, at roughly a $550 million valuation. Eleven months later that figure is 2.7x higher, with revenue up more than fivefold year over year.

The platform now runs more than two million code reviews a week for over 17,000 customers, among them Adyen, BMW, Indeed, JFrog, NVIDIA, Trivago and Campfire, and more than 150,000 open source projects use it. The company said it will put over $10 million into keeping AI code review and agent capabilities free for open source projects and maintainers over the next 12 months.

“As AI becomes critical infrastructure for the global economy, organisations will increasingly need independent governance layers that can validate software regardless of which model produced it,” Eisenstecken said. “CodeRabbit’s independent review system and deep contextual understanding position it as the control layer for agentic software development.” Kasper Sage, managing partner at BMW i Ventures, added that “the digital ambitions of automotive companies depend on highly efficient engineering teams and consistently reliable software,” calling CodeRabbit a key technology for improving code quality.

All three capabilities are available now and reach existing customers immediately; new customers get a 14-day trial. CodeRabbit recently opened a London office in Moorgate after strong pull from sophisticated European enterprises, scaling to 50 full-time employees across London and the EU, and plans to deepen that footprint while entering Japan and other key Asian markets.

The competition: writing code versus vetting it

AI code review split off from the same root as vibe coding. Once generation tools multiplied pull request volume, the cleanup became a category of its own. Whether that layer stays an independent business or gets absorbed by the tools that write the code is the open question — and so far, both are happening.

Greptile is the independent case. Rather than reading only the diff, it indexes the entire repository first and competes on bug detection rate. Founded in 2023 by three Georgia Tech graduates, it raised a $25 million Series A led by Benchmark in September 2025 at a $180 million valuation, bringing total funding to $30 million, with customers including Brex and Substack.

Graphite is the absorbed one. Known for stacked pull requests, which let developers layer dependent changes without waiting on approvals, it raised a $52 million Series B led by Accel in March 2025 and was serving over 500 companies including Shopify, Snowflake, Figma and Perplexity. That December, Anysphere — the maker of Cursor — acquired it. Terms weren’t disclosed, but the cash-and-stock deal reportedly landed well above Graphite’s $290 million valuation. Cursor also ships its own reviewer, BugBot. That deal is the clearest reason CodeRabbit keeps returning to the word independent.

Qodo takes a rules-based approach centered on ticket compliance and engineering standards. Blacksmith comes at the problem from another direction, moving from CI infrastructure — the systems that automatically build and test code — into verifying what AI writes; it raised a $45 million Series B on August 12, a tenfold valuation increase in a year. On the security side, CodeRabbit Security pushes into ground long held by static analysis and security incumbents like Snyk and Semgrep.