Putting AI into real healthcare workflows
What actually got built in practice — what was done, how it landed, what results it produced, and where we took the wrong turn. Each case aims to show checkable evidence: real product interfaces, stage metrics, and the mistakes made along the way. Everything is redacted; the focus is method, judgment and boundaries.
Intelligent Clinical Documentation and Quality Control
Embed AI in the clinician’s real workspace so generation, quality control, and sign-off reduce repetitive documentation while turning validated practice into reusable institutional knowledge.
Patient Agent and Precision Patient Operations
Move beyond digital entry points toward a continuously available health twin that understands patient state, initiates action, and completes clinical service loops under human confirmation and full audit.
From a One-Sentence Request to an AI-Native Service Desk for a Healthcare Group
Seventeen days from a three-agent concept to real production. I came away with a different understanding of AI-native enterprise software—not a model taking over everything, but a closed loop across natural language, business governance, reliable transactions, and organizational learning.