Screaming in the Cloud Conversations at the Intersection of AI and Code with Harjot Gill
Sep 4, 2025
Harjot Gill, CEO and co-founder of CodeRabbit, discusses how AI is transforming code reviews. He shares insights on building the most-installed AI app on GitHub, highlighting the inefficiencies of traditional code review processes and the challenges posed by AI-generated code. Gill explains how CodeRabbit uses innovative cloud architecture to enhance efficiency and navigate the complexities of code reviews. He also delves into balancing feedback from AI tools and the strategic advantages of using Google Cloud for startups.
AI Snips
Chapters
Transcript
Episode notes
Weekend Hack Turned Product
- Harjot built an early prototype during a weekend hack to automate code review using language models. He spun that into CodeRabbit after seeing it dramatically reduce review bottlenecks at his startup.
Why Code Reviews Fail
- Code reviews are widely broken: teams either rubber-stamp large PRs or get stuck in toxic, endless back-and-forth. Good reviews require deep contextual understanding and often take as long as writing the code.
AI Creates Review Backlog
- AI-generated code increases PR volume and complexity, shifting the burden onto senior reviewers. That creates backlog and maintenance nightmares, especially for open source projects receiving huge AI-made PRs.