Work
Decision systems for regulated, high-stakes institutions.
How It Works
Analysts run scenarios → Decision memo generated → Stakeholder review → Approval with audit trail preserved
Every decision gets a reviewable trail: what assumptions changed, which constraints bind, what trade-offs were accepted.
Healthcare Decision Systems
Context
US academic medical centers. Healthcare facility closure decisions, bed allocation, resource planning under regulatory constraints.
Decision Point
"Which facilities to close, when, and how to reallocate resources without violating access requirements."
Method
Causal inference → scenario simulation → constrained optimization
Artifact
Decision memo with constraint log, trade-off table, and governance-ready summary.
Outcome
Framework adopted for multi-site closure planning. Reduced decision cycle from months to weeks.
Proof
1st Place, POMS Healthcare Ops Best Paper Award. AHRQ R36 Dissertation Grant (PI).
AI Governance & Autonomous Systems
Context
Defense and public-sector AI deployments. Autonomous drones, decision-support systems requiring regulatory compliance.
Decision Point
"How to structure AI accountability—who approves, what constraints apply, what happens when it fails."
Method
Governance framework design → accountability mapping → audit trail architecture
Artifact
Accountability framework with decision trail documentation, constraint registry, and approval workflows.
Outcome
Framework presented to regulators and industry. Governance principles adopted in pilot programs.
Proof
Invited keynote, AI·Drone Conference (2024). AI Defense Lab opening, Daejeon.
Doogooda (Entity Flow)
Decision Intelligence Platform
Context
Enterprise clients in regulated sectors. Healthcare operations, public-sector planning, institutional risk management.
Decision Point
"Turn analytics into defensible actions—with explicit assumptions that leadership can approve or reject."
Method
Causal inference → scenario simulation → optimization → audit trail generation
Artifact
Decision memos, constraint logs, trade-off tables. Operators review scenarios; leadership signs off with preserved audit trail.
Outcome
Platform deployed with institutional clients. Decision approval time reduced 60%+ in pilot.
Proof
TIPS (Korea MSS) 2024. Invest Seoul CORE 2025. Pre-A funded.
Public Sector & Education Planning
Context
Korean provincial government. AI-driven school facility optimization, resource allocation across districts.
Decision Point
"Which facilities to consolidate, how to allocate resources across schools while meeting equity constraints."
Method
Demand forecasting → constraint mapping → scenario comparison → policy recommendation
Artifact
Policy brief with scenario comparisons, trade-off analysis, and implementation roadmap.
Outcome
Research adopted by Gyeonggi Provincial Council. Final report delivered to legislature.
Proof
Gyeonggi Provincial Council commissioned research (2024). Public sector reference.
Credentials: Harvard PhD · Yale MA · Caltech BS · Former Assistant Professor, UCL School of Management
Interested in collaboration? → Media & Speaking