PatchWave
PatchWave is a GitHub app that analyzes, fixes, and merges Dependabot pull requests. Dependabot opens a pull request for every dependency update, but it can’t tell you whether an update affects your code. So, someone on the team still has to read the changelog, check whether the breaking changes touch anything they use, and fix the build when it fails. PatchWave does that work. It reads the upstream changes, checks how your code uses the package, and links every conclusion to its source. When an update breaks the build, it writes a fix for a person to review, and it merges the updates your team’s policy allows.
At ContextBridge, I share the engineering on PatchWave with my co-founders and work across the whole product, from the AI agents and web app to the AWS infrastructure underneath. We’re an AI-native team, so we use agents throughout our own development. That let us go from the decision to build PatchWave to a public launch in 16 weeks, with evals for the AI agents running in CI within the first two weeks.
Since PatchWave runs AI agents against customer source code, keeping that code contained was a priority from day one. Each analysis runs in its own isolated environment that’s thrown away when the run ends, and the models run inside AWS with zero data retention.


Key Skills
- AI Agent Development and Evaluation
- Securely Running AI Agents on Customer Code
- AWS Architecture and Infrastructure as Code
- GitHub App Development
- Full-Stack Web Development
- Usage-Based Billing
For more information on PatchWave, visit patchwave.ai.
Get in Touch
Interested in working together? Let's connect.
Contact me