Writing
Notes from production
What actually works in AI engineering. No listicles, no hype.
Deploying LangChain Apps on AWS Lambda: A Practical Guide
2026-06-28Serverless is the cheapest way to run LLM backends at low-to-medium volume — if you avoid these traps.
AWSLangChainFastAPITutorial
RAG vs Fine-Tuning: Which One Does Your Product Actually Need?
2026-06-15Most teams reach for fine-tuning when they need retrieval. A practical decision framework with real cost numbers.
RAGFine-tuningGenAI