Show HN: Wayfinder – A reference implementation for evaluating AI applicationshttps://github.com/DivakarUngatla/wayfinder
Testing is critical to software applications — first to make sure they reliably serve the purpose they were built for, and then to make sure they stay that way as they grow, evolve and change. The same applies to AI applications as well. But testing AI applications is very different from testing classic software applications, especially because of their non-deterministic nature. It is very important for AI Engineers to develop a deep understanding of AI Evaluation to be able to build reliable AI applications and ship changes faster without breaking their existing behavior. AI Evaluation is…