Qodo raises $70M for code verification as AI coding scales

AI coding tools generate billions of lines of code each month, creating a new bottleneck in ensuring software works as intended. Qodo, a startup that develops AI agents for code review, testing, and governance, is betting that validation will define the next stage of software development.

The New York-based startup has raised $70 million in a Series B round led by Qumra Capital, bringing its total funding to $120 million. Maor Ventures, Phoenix Venture Partners, S Ventures, Square Peg, Susa Ventures, TLV Partners, Vine Ventures, Peter Welinder (OpenAI), and Clara Shih (Meta) also participated in the round.

Qodo aims to serve as a layer focused on improving the reliability of AI-generated code as enterprises accelerate the adoption of tools such as OpenClaw and Claude Code. Many people are beginning to realize that faster code output does not necessarily lead to reliable and secure software.

While most AI review tools focus on what changed, Qodo focuses on how code changes impact the overall system, taking into account an organization’s standards, historical context, and risk tolerance to help businesses manage AI-generated code with more confidence.

Itamar Friedman, who previously co-founded Visualead and led the machine vision business at Alibaba (which acquired Visualead), founded Qodo in 2022. He told TechCrunch that two key moments in his career inspired him to launch Qodo just a few months before launching ChatGPT: working at Mellanox, which was later acquired by Nvidia, and building Visualead.

At Mellanox, we worked on automating hardware verification using machine learning and realized that “system generation and system verification require completely different approaches – different tools, different ways of thinking.” Then, at Alibaba’s Damo Academy, I witnessed AI evolve into systems that can reason beyond human language. By 2021-2022, just before GPT-3.5, it has become clear that AI will generate the majority of the world’s content, especially code, reinforcing his view that fundamentally different systems will be needed to generate and verify code.

A recent study found that while 95% of developers do not fully trust AI-generated code, only 48% consistently review code before committing, highlighting the gap between awareness and practice.

“Code generation companies are primarily built around LLMs, but when you think about code quality and governance, LLMs alone are not enough,” Friedman said. “Quality is subjective. It depends on organizational standards, past decisions, and tribal knowledge. LLMs can’t fully understand that context. It’s like taking a great engineer from one company and asking them to review code at another company. They lack internal context.”

Companies like OpenAI and Anthropic are helping shape the broader AI story, including in adjacent areas like code review, but are primarily focused on building capabilities rather than end-to-end solutions, Friedman explained. There are other startups in the space, but many are still in the early stages and have yet to see widespread adoption in enterprises, the CEO said.

Qodo focuses on performance to stand out in a crowded market. The startup was recently ranked #1 on Martian’s Code Review Bench with a score of 64.3%. This is more than 10 points ahead of the next competitor and over 25 points ahead of Claude Code Review. This benchmark highlights its ability to catch tricky logic bugs and cross-file issues without overwhelming developers with noise.

Last month, the company launched Qodo 2.0, the current benchmark-leading multi-agent code review system, and introduced tools to learn each organization’s definition of code quality.

The company already partners with major companies like Nvidia, Walmart, Red Hat, Intuit, and Texas Instruments, as well as high-growth companies like Monday.com and JFrog.

“Every year there have been defining moments, from Copilot to ChatGPT to complete task automation,” Friedman says. “We are now entering a new phase: from stateless AI to stateful systems, and from intelligence to ‘artificial wisdom’. That is what Qodo was created for.”