fixa uses a voice agent to call your voice agent and an LLM to evaluate how the conversation went. here's a demo: https://youtu.be/LAW1wW6SjTo

this was initially a paid feature of our platform, but the monetization model didn't really make sense.

our customers were paying us for all the components of our agent (STT + LLM + TTS + markup) in addition to their own agent – even though many of them have preferred vendors, volume discounts, and credits.

we came to the conclusion that you should be able to use your own API keys to test your own voice agents.

e2e testing for web apps is done using free and open source packages (selenium, puppeteer), we think the same should be true for voice.

running tests with fixa will always be free. we make money if you decide to use our cloud platform to visualize your test results.

fixa is still in the early stages, so we would appreciate any and all feedback!