- Published on
- •7 min read
Part 4 covers patterns 07–15 — orchestration, safety, memory, RAG, and event-driven agents — plus how to move from runnable examples to production architecture.
Kamrul Hasan — Senior Data and AI Consultant
I build scalable data platforms and ship practical AI systems — from ingestion and MLOps to agentic workflows. Recently moved to a new team; writing here about what I learn along the way.
About Me
Senior Data and AI Consultant with deep experience in data engineering, MLOps, and agentic AI. I build trustworthy data foundations and help teams deliver end-to-end outcomes.
Read my full story →What I Do
🔧Data Engineering & Platforms
🤖Agentic AI, ML & Deep Learning Solutions
📊Data Governance & Strategy
🧑💼Consulting & Mentorship
Latest
A blog about programming, data and web development.
- Published on
- •6 min read
Part 3 covers patterns 01–06 — ReAct, Tool Use, Planning, Prompt Chaining, Routing, and Parallelization — with LangGraph flow diagrams and runnable commands for each.- Published on
- •6 min read
Part 2 explains the three shared scenarios (IT helpdesk, e-commerce, ML pipeline) and the shared runtime that lets every pattern example run with the same CLI, tools, and LangGraph helpers.- Published on
- •6 min read
Part 1 introduces agentic-ai-overview — a hands-on catalog of 15 LangGraph design patterns with three shared scenarios, and explains how to learn agentic AI by comparing patterns on the same domain.- Published on
- •8 min read
Part 4 covers hexagonal architecture, the agent memory model, the background worker, Docker deployment, and the live agent explorer for learning multi-agent systems.
Subscribe to the newsletter