codex is great for like a one-time/overview analysis on a handful of transcripts. we usually serve to companies where the volume is >10k messages & continuous ingestions + with claude/codex it messed up this + metadata linking of the user like what plan are they on, when is it expiring, etc.
although we had a few customers who come to us after running this for a while so at smaller volume it does work well.
To be fair if you have some background in data science (I think the main thing is being able to verify its not showing nonsense) it shouldn’t be too hard to build a bespoke solution.