writing

Articles, tutorials, and thoughts on AI systems, MLOps, and production machine learning.

Apr 01, 2026 Why Most Agentic AI Systems Fail in Production (And How to Fix That)
Agentic AI demos look great. Production deployments rarely do. Here's the engineering discipline that bridges the gap.
Mar 20, 2026 RAG Retrieval Metrics You Should Actually Be Tracking
Hit Rate@K, MRR, NDCG — what they measure, when each one matters, and how to implement them for your RAG system.
Feb 15, 2026 MLOps Monitoring: The Tools Actually Worth Using in 2026
A practitioner's comparison of Evidently, Arize Phoenix, Azure ML Monitor, and LangSmith for production ML and LLM monitoring — based on real deployments.