I like building AI systems that are safe, convenient, and reliable. What matters most to me is making complex systems feel dependable and genuinely useful in real-world workflows.
current focus: AI infrastructures and agent orchestration, Reinforcement learning, and Governance, Risk, and Compliance for AI agents.
An SDK and runtime infrastructure for governing AI workflows with context-aware guardrails. We capture trajectory data from orchestrator decisions, agent actions, tool calls, policy outcomes, and HITL interventions to support safe, continuous agent improvement.
Researched chunking a trajectory prediction model into stages and ran them in a distributed pipeline across simulated edge devices to evaluate latency and throughput.
Built a Java simulator to help visualize how CISC (CPUs that execute complex instructions) processor executes instructions and how it uses memory and cache.
I spent my time here researching, architecting, and tuning computer vision models for human pose estimation pipelines, building POCs, addressing edge cases that are often overlooked, and translating human biomechanical metrics into ML estimations used to conduct injury risk assessments and improve workplace safety. More details
Skills I care deeply about core CS fundamentals including principles of programming languages, computer systems design, computer networks, cloud computing, and database systems. To me, those foundations matter more than being tied to a small set of specific languages or frameworks.