Experiments

Prototypes, side projects, and technical curiosities across software, AI, web platforms, and product ideas. A place for rough edges, half-formed questions, practical tests, and things I’m building, breaking, refining, and figuring out in public.

ARTIFICIAL INTELLIGENCE EXPERIMENTS

Reasoning About AI Reasoning: Letting the LLM Cook

A practical essay on why LLMs may outperform declarative AI workflows when given the right tools, context, and room to reason at runtime. Explores MCP, dynamic skill fetching, right-sized agents, recursive orchestration, and the guardrails needed when we let AI handle higher levels of abstraction.

Read