SoC-level formal and simulation verification of Flash, XSPI, PLL and RTC subsystems, plus Python and Tcl automation tooling. Presented the Maestro XML Diff tool at DVCon 2025.
About me
I’m a Design Verification Engineer at Texas Instruments and an MS student in Computer Science at Northeastern. I’m applying to PhD programs for Fall 2027 to study multi-agent learning, specifically how failures arise and spread in teams of learning agents and how to detect them early.
My earlier research was in hardware security. In the FICS Lab at the University of Florida, I worked on privacy-preserving chip design tools built on garbled circuits (ICCAD 2022), side-channel protection for neural-network inference on FPGAs (SPACE 2022), and IC metering protocols that resist fault injection (SRC TECHCON 2022).
Since 2026 my work has moved to machine learning. Comparing MADDPG with independent DDPG on a multi-agent deception task, I found that with three agents sharing one critic, the team suppressed the adversary but failed to cover its target, a likely case of reward hacking. I have also probed what a DreamerV3 world model encodes about small, task-critical objects.
At TI, I verify SoC subsystems with formal and simulation-based methods. I want to bring the same habits, explicit failure modes and checkable guarantees, to multi-agent systems.
Research interests
I’m interested in multi-agent systems: how agents learn, coordinate, and fail, and how to make them robust, reliable, and safe.
Multi‑Agent Learning & Coordination
How teams of learning agents cooperate, compete, and coordinate under partial information, in multi-agent RL and LLM agent systems.
Fault Detection & Forecasting
Detecting errors early and forecasting how they spread through a team of agents, so systems can recover before one agent’s mistake becomes a team failure.
Safe & Secure Agent Systems
Keeping agent systems safe and secure, from guarantees in safety-critical settings to agents that act on people’s behalf and keep them in control.
Publications
Details & talks →-
Active IC Metering Protocol Security with Oblivious Transfer
Experience
Full-chip AXI interconnect tests between the SoC and Southbridge for the Xbox platform, with coverage analysis and debugging.
Hardware security research leading to three 2022 papers: garbled-circuit chip design tools (ICCAD), side-channel-protected neural networks (SPACE) and secure IC metering (SRC TECHCON).