Qodo raises $70M to review AI-generated code before it ships

In short: Qodo raised $70 million to build software that reviews AI-generated code and flags problems before teams release it.

What happened

Qodo is betting that the hard part of AI coding is not writing code fast, it is making sure the code actually works. The company sells a code review tool that checks changes for bugs, inconsistent behavior, and rule violations before they get merged into a product.

Code review is the step where someone reads new code and looks for mistakes. Qodo tries to automate more of that work using AI, while also using the full “context” of a company’s existing code, meaning it looks across the whole codebase instead of only the new lines being added (like checking a whole cookbook, not just the new recipe).

The company says its system can run multi-step review workflows, such as finding issues, suggesting tests, and validating a change before it is merged. It also generates and updates rules based on past reviews and common patterns, so teams can enforce their standards without relying only on a static checklist.

Qodo can be used in common developer tools and code hosting services, including GitHub and GitLab. TechCrunch reports that Qodo has been featured in an NVIDIA GTC keynote and that NVIDIA uses it for code search at scale. Qodo also points to a developer survey showing many engineers are not confident shipping AI-written code without human review.

Why it matters

As more software is written with AI, more mistakes can slip through simply because there is more code to check. Tools like Qodo are trying to act like an extra set of eyes, so companies can move faster without ignoring safety, reliability, and internal rules.

Source: TechCrunch AI