This portfolio sample shows how I evaluate AI agents and automated workflows to find reliability problems, broken logic, and weak edge-case handling. I review whether the system completes the intended task, uses tools correctly, follows instructions, handles ambiguity, and produces an output that can actually be trusted.

The work focuses on identifying failure points before they become real user problems, including incomplete task execution, incorrect assumptions, workflow dead ends, hallucinations, and verification gaps. This is a portfolio demonstration of my AI agent evaluation and red-teaming work. No confidential client information is included.

Thorough breakdown! Robust edge-case handling and reliability checks are definitely the trickiest part of building reliable AI agents 🤖📊

Finally posting this, back in May we worked on a trial project for Boona 🚿

Boona is a tandem shower brand, which was our first time working on a brand like this but it was soooooooo fun!

The brief was open enough that we had to dig, not just the Instagram feed but the website, the product itself, what it actually feels like to share that experience with someone. Everything kept coming back to the same thing: intimacy, closeness, two people in a moment that's usually just one.

We didn't end up working together but the founder's feedback made it worth sharing 🤩

I don't usually share my trial work but I'm thinking it's about time to do it, what do you think?

Glad you decided to share it, trial work like this is often more interesting than the polished final stuff because you can actually see the thinking behind it. 'Goosebumps before the water even hits' is such a strong line, did that come from the brief or from digging into the product yourself?