Today's open source release: QVAL
A desktop app for macOS that runs LLM evaluations on support tickets according to whatever rules and schema you provide. It's a great tool for testing your LLM rules and prompts to refine them before putting them into production. Run LLM evals and human evals, compare them against each other, and merge with your teammates' evals to get a group consensus.
Free to download and use. Feedback and contributions are welcome!