I build decision-making systems that hold up under real-world uncertainty.
Ph.D. candidate in ECE at Vanderbilt University. I combine optimization (MILP, MPC), reinforcement learning, and multi-agent coordination to turn messy, uncertain data into robust, deployable policies for EV-grid and mobility systems, from peer-reviewed theory to field-validated tools.
Vehicle-to-Building (V2B)
Online control under uncertain EV arrivals, building loads, and user behavior, balancing user utility with grid objectives.
Sequential Decisions Under Uncertainty
Scenario generation, stochastic optimization, MCTS, and rolling-horizon MPC for real-time cyber-physical control.
Multi-Agent Coordination
Strategy-aware incentive mechanisms and negotiation frameworks that align individual agents with system-level goals.
Digital Twins & Simulation
Open-source toolchains for transit and mobility-energy planning: E-Transit-Bench, OPTIMUS, MoveOD.
Hierarchical EV-Grid Control
MILP + Multi-Agent RL coordinating 50+ distributed energy assets; >10% cost reduction on field data.
Electric Bus Scheduling
Stackelberg game for charging schedules in Chattanooga; −14% cost, −10 tons GHG emissions.
MoveOD Pipeline
Fuses Census + traffic data into calibrated origin-destination matrices. 2nd Place, INFORMS 2025.