RAG Evaluation Metrics: How to Test a RAG Chatbot Before You Ship ItHow to evaluate a retrieval-augmented chatbot: separate retrieval from generation, build a real test set, choose checks you can repeat, and report results honestly.6 min read·boilershelf.hashnode.dev17m ago
AI SaaS Architecture: How to Structure a FastAPI and Next.js ProductThe architecture decisions that matter when you build an AI SaaS: auth, tenancy, an LLM gateway, background jobs, usage metering and billing.6 min read·boilershelf.hashnode.dev12h ago
