Qodo Raises $70 Million to Secure AI Code

With the rise of AI-based coding tools that generate billions of lines of code each month, a new challenge arises: ensuring that this software functions as intended. Qodo, an innovative startup based in New York, is positioning itself to meet this need by developing AI agents specialized in code review, testing, and governance. The company recently raised $70 million in a Series B funding round led by Qumra Capital. This funding round brings Qodo's total financing to $120 million, with notable contributions from Maor Ventures, Phoenix Venture Partners, S Ventures, Square Peg, Susa Ventures, TLV Partners, Vine Ventures, as well as Peter Welender from OpenAI and Clara Shih from Meta.

Qodo focuses on enhancing trust in AI-generated code, a crucial issue as companies increasingly adopt tools like OpenClaw and Claude Code. However, the speed of code production does not necessarily guarantee its reliability or security. Qodo stands out for its ability to assess how code changes impact entire systems, taking into account organizational standards, historical context, and risk tolerance. This approach enables companies to better manage AI-generated code.

Founded in 2022 by Itamar Friedman, a former co-founder of Visualead and head of computer vision at Alibaba, Qodo leverages Friedman’s experience in automating hardware verification and the evolution of AI towards systems capable of understanding human language. Friedman was inspired by his experiences at Mellanox, acquired by Nvidia, and at Alibaba's Damo Academy, to create Qodo just before the launch of ChatGPT. At Mellanox, he worked on automating hardware verification using machine learning, realizing that generating and verifying systems require very different approaches.

A recent survey reveals that 95% of developers do not fully trust AI-generated code, yet only 48% systematically review it before committing. This highlights a gap between awareness and practice. Friedman emphasizes that language models, or LLMs, are not sufficient on their own to ensure code quality and governance, as they cannot fully understand the organizational context. "Quality is subjective," he explains. "It depends on organizational standards, past decisions, and tribal knowledge. An LLM cannot fully grasp this context. It's like taking an excellent engineer from one company and asking them to review code in another — they lack the internal context."

Companies like OpenAI and Anthropic play a role in shaping the broader narrative of AI, including in adjacent areas like code review, but they primarily focus on building features rather than end-to-end solutions, according to Friedman. While there are other startups in this space, many are still in the early stages and have not yet achieved widespread adoption by enterprises, noted the CEO.

Qodo is banking on performance to stand out in a saturated market. The startup was recently ranked number one on the Martian Code Review Bench, with a score of 64.3% — over 10 points ahead of the next competitor and 25 points ahead of Claude Code Review. This benchmark highlights its ability to detect complex logical bugs and inter-file issues without overwhelming developers with noise.

In the past month, it launched Qodo 2.0, a multi-agent code review system that currently leads the benchmarks, and introduced tools that learn each organization's definition of code quality. The company is already collaborating with major firms such as NVIDIA, Walmart, Red Hat, Intuit, and Texas Instruments, as well as high-growth companies like Monday.com and JFrog.

Friedman concluded: "Every year has had a pivotal moment — from Copilot to ChatGPT to full task automation. Now, we are entering a new phase: moving from stateless AIs to stateful systems — from intelligence to 'artificial wisdom.' That’s what Qodo was designed for."