yep! we&#x27;ve seen that a lot of voice agent developers are manually testing their agents by making calls themselves. some of these calls can get up to 15 minutes long, which makes it really hard and time-consuming to iterate. we created this as a way to help with that iteration process :)

Super interesting, I imagine for production use cases of voice agents; Some form of testing and observability into voice agents will be necessary; to fix all the various edge cases

HAHAHA glad we could help you live out your dreams of being called by Steve Wozniak

great work guys! love that it&#x27;s open source (gave myself a funny call from Steve Wozniak) - excited to see where this&#x27;ll go

hey! this is oliver and jonathan and today, we’re launching fixa — a free, open source package to test voice agents.fixa uses a voice agent to call your voice agent and an LLM to evaluate how the conversation went. here&#x27;s a demo: <a href="https:&#x2F;&#x2F;youtu.be&#x2F;LAW1wW6SjTo" rel="nofollow">https:&#x2F;&#x2F;youtu.be&#x2F;LAW1wW6SjTo</a>this was initially a paid feature of our platform, but the monetization model didn&#x27;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!

Show HN: Fixa – an open source Python package for testing voice agents