Harjot Gill was leading FluxNinja, an observability startup he co-founded a few years following the sale of his first company, Netsil, to Nutanix in 2018, when he began to notice an interesting development.
“We had a distributed engineering team that started using AI-driven coding tools like GitHub Copilot,” Gill explained to TechCrunch. “As we watched this trend take off, it became obvious to me that it would eventually create new bottlenecks in the code review process.”
In early 2023, Gill launched CodeRabbit, a platform for AI-driven code review, which later took over FlexNinja.
His forecast proved accurate: developers have widely adopted AI assistants for coding, but the resulting code frequently contains flaws, requiring engineers to devote significant effort to corrections.
CodeRabbit is able to detect some of these mistakes. The company has been expanding by 20% each month and, according to Gill, has surpassed $15 million in annual recurring revenue (ARR).
Investors are enthusiastic about CodeRabbit’s progress. On Tuesday, the company revealed it secured $60 million in Series B funding, giving it a $550 million valuation. This round, which increased the startup’s total funds raised to $88 million, was led by Scale Venture Partners, with participation from NVentures, Nvidia’s venture capital branch, and previous investor CRV.
CodeRabbit is enabling companies such as Chegg, Groupon, and Mercury—as well as more than 8,000 other organizations—to streamline the often tedious process of code review, a task made even more challenging by the proliferation of AI-generated code.
Gill stated that CodeRabbit’s understanding of a company’s codebase allows it to identify bugs and offer suggestions, effectively functioning as a team member. He also mentioned that businesses using CodeRabbit can reduce their human code reviewers by 50%.
Like much of the AI sector, CodeRabbit faces competition. Other startups in this space include Graphite, which recently landed $52 million in Series B funding led by Accel, and Greptile, which is reportedly negotiating a $30 million Series A with Benchmark.
Although major AI coding tools such as Anthropic’s Claude Code and Cursor provide their own AI-based code review features, Gill is confident that, over time, clients will gravitate toward a specialized, standalone solution. “CodeRabbit offers a much deeper and broader technical approach compared to all-in-one packages,” he said.
It’s still uncertain whether Gill’s prediction will ultimately be proven right. For now, though, thousands of engineers are willing to pay CodeRabbit $30 a month.
Despite the rise of AI-powered code review platforms like CodeRabbit, current AI tools are not yet completely reliable for fixing bugs and “unusable” code generated by AI systems. As a result, a new job title has emerged within companies: the vibe code cleanup specialist.
