I troubleshoot AI agents.
So I built one to troubleshoot — for you.
Instead of a résumé, I built a working sample of the exact craft this role is about: reviewing conversations, diagnosing what went wrong, and refining the prompts that make agents reliable. Click around — every tool here works.
What I bring
Systematic debugging, precise guard-railed prompt writing, systems thinking, pattern-spotting across hundreds of conversations, and clear writing that makes agents behave.
Why I built this
I wanted to show the daily review → diagnose → fix loop in action, not just claim it on a one-page résumé.
How to use this
Open the Conversation Troubleshooter to watch me fix a failing chat, then try the Conversation Evaluator to see live reasoning over buyer messages.
I'd love to grow with Lahzo.
This role is the exact intersection of writing, systems thinking, and pattern-spotting I've been working toward. I'm ready to plug into the team, learn the stack, and start shipping real prompt fixes from week one — and grow from AI Designer I into whatever the team needs next.
Illustrative sample, not affiliated with Lahzo; all conversations are fictional.