Thoughts, tutorials and insights.
I write about AI, full-stack development, automation, and building products that solve real problems. Tutorials, guides, and lessons from my journey.
How I Built a Production RAG System Answering 100K+ Documents at 94% Accuracy
A first-hand case study: the architecture, retrieval tuning, and lessons from shipping a RAG system that answers over 100K documents at 94% accuracy.
How Much Does It Cost to Build a RAG System in 2026? A Real Line-Item Breakdown
What a production RAG system actually costs in 2026: the real cost drivers, ranges by project type, and where teams overspend.
RAG vs Fine-Tuning: A Decision Framework (With Cost Math)
When to reach for retrieval, when fine-tuning earns its place, and why most small teams shouldn't fine-tune first.
Advanced RAG: The Chunking, Hybrid Search & Reranking That Moved My Accuracy to 94%
The chunking, hybrid search, and reranking tactics that take a RAG system from a naive baseline to 94% retrieval accuracy.
Qdrant vs Pinecone vs Weaviate: Choosing a Vector DB for Production RAG
An honest, production-grade comparison of Qdrant, Pinecone, and Weaviate: hosting, pricing, filtering, scaling, and when to self-host.
How to Actually Evaluate a RAG System (RAGAS, Metrics & Production Monitoring)
How to measure whether a RAG system actually works: retrieval vs generation metrics, RAGAS, golden eval sets, and monitoring drift in production.