I explore the architecture of autonomous agent systems. My interests lie in the infrastructure that makes the web readable and actionable for AI.
- **Agentic Design Patterns:** How agents discover tools, verify data, and execute tasks reliably (MCP, Function Calling).
- **Data Curation & Evals:** Building high-fidelity datasets that serve as ground truth for domain-specific models.
- **Human-AI Interaction:** Designing interfaces where human intent translates seamlessly into agentic action.
Currently focused on transforming unstructured knowledge into transaction-ready APIs for the next generation of LLMs.