We just launched AI that writes your bug reports and today, we’re sharing the story with the team that built it! We’ll get into why we use Smaller Language Models instead of the larger LLMs, moved to an opt-in user experience, and all the user feedback that shaped what the feature is today.

Let's get into it!
 
(00:00) Intro
(01:14) Why we changed the UX from opt-out to opt-in
(02:35) How early access helped us fine-tune AI (not everything people Jam is a bug!)
(03:53) Setting our quality bar to ship AI & how to measure it
(06:02) Our biggest early access breakthroughs: auto-generated repro steps & short titles
(08:55) Decisions we made to prioritize AI speed & accuracy
(12:15) The story of how Jam AI was engineered from a van, Greece & sunny beaches

Check out Jam AI at: https://jam.dev/ai

Subscribe to Building Jam on YouTube, Spotify, and Apple Podcasts.


New episodes drop every Friday at 10AM ET. See you there!

Podden och tillhörande omslagsbild på den här sidan tillhör Jam.dev. Innehållet i podden är skapat av Jam.dev och inte av, eller tillsammans med, Poddtoppen.