Hey everyone! Mukund and Ari here, and we wanted to show you folks what we've built. Our platform ingests your codebase, develops a semantic understanding of classes + functions, and generates unit tests via LLMs.

When a developer posts a PR with an untested function, we check out their branch, write and refine test cases, and post them back as a PR comment. This ensures you get to 100% code coverage with minimal dev lift.

How is this different from TestPilot?

- We support Swift!

- Unlike generic solutions like TestPilot, we have a deep understanding of your codebase, so our tests utilize actual semantic structure. This means higher-quality tests that don’t just examine the internals of your function in a vacuum. Because we understand your code, we’re able to test it in the context of related modules.

- These generated tests don’t just live in your IDE! Because they’re posted onto GitHub, other engineers are able to review them as well.

We’re looking for alpha testers. If you’re working on a Swift side project, we'd love to work with you so we can build better unit tests!

Please reach out mukund.tibrewala at gmail dot com