I design and evaluate applied AI systems with a focus on agent workflows, tool use, guardrails, observability, structured data, and human oversight.
My work combines hands-on AI engineering with product and organizational thinking: whether a system can operate reliably, explainably, and usefully in the environment around it.
Implemented systems, prototypes, and concepts are labeled clearly. Each project shows its boundaries and next step.
Agentic Analytics Lab
A Python analytics-agent baseline on synthetic operational data, with dataset-scoped read-only ClickHouse tools, semantic metric guards, evaluation infrastructure, automated regression coverage, CI, and CodeQL. A reviewed oracle-metadata routed execution experiment now tests when specialized execution is worth the added control-plane complexity.
HomeOps Agent
Agent tools, explicit tool contracts, approval gates, failure handling, auditability, and deterministic scenarios. Current device integrations are simulated for reproducible testing.
Closing the Loop
Measurement design for activation, task completion, follow-through, and friction signals, with workflow reasoning and responsible-AI considerations. Concept and analytics case study, not production software.
Useful AI needs a baseline, a boundary, and a way to learn from failure.
Start with a baseline, define success, then justify additional architecture with evidence.
Permissions, tool access, semantic rules, approval gates, and human oversight belong in the architecture.
Incorrect answers, tool failures, latency, and edge cases should be measured and documented.
A technically interesting model is not automatically a useful system.
My path into applied AI is interdisciplinary. Psychology taught me to pay attention to behavior, trust, and how people make decisions. Business and operations work taught me to understand the systems surrounding a technology. Applied AI pushed me deeper into building, testing, debugging, and evaluating the technology itself.
Today I’m especially interested in the gap between an impressive AI demo and a system someone could responsibly depend on.
Alongside my engineering work, I serve as an Applied AI Programs & Operations Coordinator at the Paul English Applied Artificial Intelligence Institute at UMass Boston, where I work around AI programs, learning experiences, workflows, operations, and adoption. I am also completing an MBA at UMass Boston.
I’m interested in Applied AI Engineer, AI Engineer, AI Solutions Engineer, and AI Automation Engineer roles, especially work involving agents, evaluation, enterprise workflows, knowledge-work automation, analytics, and human oversight.
Also in progress: Hermes Meeting → Action ↗