Projects
Engram→
Hybrid persistent memory layer for AI agents combining knowledge graphs, vector search, and temporal versioning. Enables agents to store, retrieve, and reason over structured and unstructured memories across sessions.
TrajAI→
Open-source testing framework for AI agents. Mock tools and assert on agent behavior rather than raw outputs — filling the gap that standard unit testing frameworks leave for agentic systems.
MCP Debug Server→
MCP server that enables Claude Code to interactively debug Python and Node.js applications — set breakpoints, step through code, inspect variables, and evaluate expressions directly from a Claude conversation.
BlinkMonitor→
macOS menubar app that uses real-time facial landmark detection to track blink rate and automatically dims the display when prolonged eye strain is detected.
Talon→
Adversarial red-teaming framework for LLM agents. Simulates multi-turn attacks to detect policy violations, jailbreaks, and unsafe behaviors. Deployable as a standalone library or a reusable GitHub Action for CI/CD.
ResumeOptimizer→
Full-stack AI resume tailoring platform. Paste a job description and your resume; AI suggests targeted improvements and rewrites. Includes database persistence for tracking multiple resume versions.
Event-Guided Video Frame Interpolation→
Neuromorphic vision research project (UW CSE 493). A dual-encoder U-Net fuses low-res RGB frames with synthetic event camera voxel streams to reconstruct sharp, high-framerate 4K video — achieving temporal resolution far beyond what RGB alone can produce.