About

I’m Nick Pizzo, a software development leader focused on applied AI engineering: turning ambiguous product and workflow problems into eval-ready LLM systems. I use this site to show how I think through retrieval, agent workflows, observability, cost and latency controls, and the production boundaries that separate useful AI applications from neat demos.

What I’m Building Toward

• Production-grade LLM applications on top of foundation models
• AI prototypes that make the launch boundary explicit: what the model decides, what stays deterministic, and how failures become regression checks
• RAG and search systems that are evaluated instead of vibe-checked
• Agentic workflows where the model’s actual decisions are explicit
• Cost and latency notes with concrete numbers, not hand-waving
• Postmortems that explain what broke and what changed afterward

Contact

• GitHub: github.com/nickpizzo
• Email: nick.pizzo@gmail.com