former data scientist pivoting fully into applied ai engineering
did the notebooks and dashboards life for years. now I want to ship real products on top of these models. excited to learn the engineering side from people who actually do it.
been saying this for months and nobody listened
anyone got a minimal example of this?
the comments here are better than most blog posts
what is the failure mode when the tool call times out?
we built something close to this, happy to compare notes
do you have a repo or gist? would love to poke at it
ok now do the version that handles errors
this is a really clean mental model, thanks
great writeup, bookmarked
more posts like this please
i keep seeing people recommend this, glad someone wrote it down
what model were you running for this?
the eval first mindset is underrated, nice to see it here
did you compare against the obvious baseline?
hard disagree honestly, in my testing it went the other way
i would be careful recommending this to beginners
appreciate you sharing the failures too, not just the wins
appreciate you sharing the failures too, not just the wins