Innovation
Designing what comes next
Concept and exploratory work. These are explorations of how decision intelligence could work — not deployed commercial products.
Pharmacovigilance Decision Intelligence
An exploration of how safety signals could move from detection to human decision with evidence and explanation attached at every step.
CareSense Copilot
A concept for ambient clinical capture where the AI summary is always a draft awaiting clinician validation.
NeuroScan AI
An exploration of imaging AI where classification is presented with confidence and explanation, and clinical review is structurally required.
Marketing Optimization / Mia
Exploratory work on conversational and scenario-driven marketing optimization, where recommendations remain proposals for a human to accept.
Writing
Thinking in public
Articles and essays on medical AI accountability, agentic interfaces and AI-assisted design practice.
Trust & accountability in medical AI
Clinical AI only earns adoption when evidence, confidence and accountability travel with the output. Most of my writing argues for structural human validation rather than optional review.
Agentic UX and its loopholes
Agents fail quietly: unclear delegation boundaries, invisible reasoning, no rollback. I write about the control surfaces agentic products still lack.
Design as a prompting discipline
Frameworks for using generative models as a design collaborator — adaptive fidelity, structured prompts and repeatable workflows rather than one-off outputs.
