Topic
Tutorials
All posts
Newest first- Nov 10, 2026 JPEG from Zero to Hero: How JPEG Compression Works, Its Variants, and an Encoder You Build in Python How JPEG compression works, told through one photo, then its variants up to JPEG AI and an encoder you build in Python. 40 min
- Nov 8, 2026 Shipping your agent How to deploy an AI agent: a FastAPI service with SSE streaming, timeouts, a shared checkpointer, Docker, tracing and a cost per… 14 min
- Nov 8, 2026 Securing an agent AI agent security in practice: session identity, authorisation in tools, prompt injection defences, PII masking, limits and audit logs. 15 min
- Nov 8, 2026 Testing an agent you can’t predict AI agent evaluation in practice: fixed test cases, answer and tool trajectory checks, an LLM-as-judge rubric and a pytest gate in CI. 14 min
- Nov 8, 2026 Many agents, or one good one? Should you go multi-agent? A LangChain supervisor with two specialists against one agent on 30 questions, plus where a Deep Agent fits. 11 min
- Nov 8, 2026 Human in the loop Add human in the loop approval to an AI agent: pause with LangGraph interrupt(), wait for a manager, and resume even after… 12 min
- Nov 8, 2026 When the loop isn’t enough: graphs Build a LangGraph workflow for leave booking: code owns the steps and rules, the model reads language, and an agent runs as… 13 min
- Nov 8, 2026 RAG as a tool, not a pipeline Build agentic RAG with LangChain: index a handbook, give the agent a search tool that cites sections, and let it say it… 10 min
- Nov 8, 2026 Give your agent a memory LangChain agent memory in practice: a checkpointer per conversation, a store for lasting preferences, and summaries for long chats. 11 min
- Nov 8, 2026 Tools that don’t break your agent Build LangChain tools an agent can recover from: Pydantic validation, errors with hints, idempotent writes and tools from an MCP server. 13 min