agents-towards-production

Why we included this project

This is one of the most practical hands-on learning paths we have found for teams that need to get a GenAI agent out of the notebook and into a deployed service. Rather than a single framework walkthrough, it is a set of runnable tutorials, each covering one building block of a real agent stack: stateful workflows, vector memory, RAG, web search, Docker and FastAPI deployment, security guardrails, GPU scaling, multi-agent coordination, observability, and evaluation. Every tutorial lives in its own folder with ready-to-run notebooks or code, so a developer can jump straight to the piece they are stuck on and work through it in minutes. It is especially useful for engineers who know the basics of LLM APIs but have not yet dealt with the operational concerns that separate a demo from a deployed service. Because the material follows the full lifecycle rather than one vendor's tooling, it also works well as a shared reference for a small team standardizing on how they build and ship agents.