AI

Reranking for two-stage RAG retrieval, from cross-encoders to TypeSafe's Jev: how it works, its limits, a 90-query evaluation, agent integration, and a prompt-injection guard.

From Copilot and Cursor to CLI agents, I share how I delegate tasks, verify AI-generated code, and keep engineering judgment in the loop.

Learn how Docker Sandboxes isolate AI coding agents with microVMs. Compare containers and sandboxes, configure network policies, and run Claude Code safely.

This article explains why modern LLM / AI chat apps often choose SSE (Server‑Sent Events) over WebSocket, comparing protocol traits, browser support, scalability, and showing end‑to‑end code examples with Hono, Zustand, and the Vercel AI SDK. It’s written for frontend and full‑stack developers building streaming AI chat experiences and deciding when to use SSE or WebSocket in real‑world projects.

Master vector search with pgvector in PostgreSQL. Covers embedding fundamentals, similarity metrics, IVFFlat vs HNSW indexes, database selection, and complete Bun/TypeScript implementation for AI-powered semantic search applications.