Testing and evaluating Copilot agents
A production-readiness case for agent testing, release gates, and review loops.
Case Studies
A curated path through applied AI, evaluation, data product, product systems, and focus-tool work for hiring managers, collaborators, and applied AI/product audiences.
How to read these
How to read these
Each case-study path highlights a different kind of judgment: production AI reliability, data product design, workflow measurement, product value, or focused user experience.
What they show
What they show
The common thread is practical systems thinking: define the user, constrain the workflow, make outputs inspectable, design for review, and connect technical work to adoption.
For collaborators
For collaborators
Start with AI evaluation if you care about shipping reliable agents, data products if you care about decision systems, and Focus Room if you want to see product interaction craft.
Related writing
Relevant articles and artifacts
AI Agent Evaluation Launch Checklist
A lead-magnet style product artifact for operational AI quality.
Healthcare AI analytics
Applied AI and analytics framing for healthcare operations.
A/B test engineering guide
Experiment infrastructure and measurement systems for product teams.
Focus Room
A product prototype with interaction design, ambient focus, and a calmer work-start ritual.
Projects overview
A broader index of public project and portfolio themes.