Skip to content
Chiranjib AI

ઘરRoadmap

Free lead magnet

AI Agents Learning Roadmap

A reading order Chiranjib uses when he explains agents to students and founders. Each link goes to someone else's official docs, free course, or essay. This page does not copy their lessons.

The searchable directory of free frameworks, models, courses and tools is on Learn. This page is the shorter reading order. Links were checked on 3 November 2026. If a vendor moves a doc, the stage still tells you what to look up. The taught version of this path, in your language and in Odia, is on the course waitlist. It is not recorded yet.

Stage 01

Foundations of LLMs

What a language model is doing, before any agent framework. Enough to tell a demo apart from a system.

Stage 02

Prompting

How you ask, what you put in the context window, and how you check the answer. Official guides first.

Stage 03

RAG

Retrieval augmented generation: answering from your own documents instead of from the model's memory.

Stage 04

Tool use and function calling

The jump from a chatbot to something that can look something up, write a row, or send a message.

Stage 05

Agent frameworks

Pick one stack and build a small agent. The framework is the easy part. State, tools and failure handling are the work.

Stage 06

Evals

How you tell whether the agent is actually working, on examples you wrote down, before a customer finds out.

Stage 07

Deployment

Put it where it can run tomorrow: logs, secrets, a host, and a way to turn it off.