Small tools for real problems. A living portfolio documenting how I frame messy problems, design product logic, build working prototypes with AI support, and iterate through real use.
Not a code portfolio. A product judgment portfolio.
PM job search generates too many daily listings. How do you filter signal from noise — and score fit — without spending 45 minutes per day?
"Explainability matters more than automation. The system should show why a role is worth applying for, not just produce a score."
Australian parking signs combine time, day, zone and permit logic. The hardest part isn't reading the sign — it's knowing what applies right now.
"The product risk is not OCR accuracy alone. It needs time, location, confidence, and a clear safety boundary."
Building with AI creates scattered decisions, prompts, screenshots and iterations. How do you turn an ongoing build process into structured product evidence?
"The portfolio is the product — not just the outputs. It should capture product judgment while the work is still unfolding."
Clarify the user, pain point, context, and what a useful outcome should look like.
Turn the problem into product logic: inputs, outputs, rules, trade-offs, and edge cases.
Use AI-assisted tools to prototype, test, debug, and refine while keeping product direction and trade-offs explicit.
Document decisions, limitations, and what should be improved in the next iteration.