Dominick Caponi·Aug 18Coding is Solved… Now What?If I had to spend time and energy getting good at something, here’s what I’d focus on… it’s not leetcode
Dominick Caponi·Jul 22When it Comes to Evals — LLMs Aren’t the Only Tool in the ToolboxOld fashioned machine learning is all you need
Dominick Caponi·May 12If Your Agent Regularly Eats Junk Food How Can it Perform Well?A gentle introduction to RAG context precision and recall for non data science nerds
Dominick Caponi·May 7Why Your AI Needs A Headlight Before A Full DashboardA cheap simple evaluation technique to identify glaring issues with your agentic system without vibes
Dominick Caponi·May 46 Enterprise Knowledge Base Quality Signals for AI AgentsIs your enterprise knowledge base really RAG ready?
Dominick Caponi·May 1When Your AI Lacks ConsistencyDo we just accept “AI is Non-deterministic” or can we do better?
Dominick Caponi·Apr 17What We Learned Teaching a Machine to Pick NBA WinnersThere’s a line from Casino that lives in my head rent-free: Sam “Ace” Rothstein, played by Robert De Niro, explains that he was “the best…
Dominick Caponi·Apr 8I built a grocery item classifier two ways: one LLM call that knows the entire taxonomy, and a…The results don’t tell a simple “one is better” story. They tell a story about engineering trade-offs determinism, maintainability, and…
Dominick Caponi·Apr 7The Reliability Stack: Building Evals That Actually Work for Agentic AIIn the last post, we established that evals are the #1 skill for forward deployed engineers. That wasn’t hyperbole it’s the “one weird…