Back to labLearn AIShovon Saha

one build ยท three tracks

Learn AI. See how it works.

Start with a practical guide, explore an interactive lesson, or build an agent step by step. Each track pairs examples with limits and ways to check the result.

6 guides

AI by profession

Everyday users, writers, teachers, finance, law and health. Scenarios, risks and verification for your actual job.

Browse guides
6 levels

AI, levelled up

The plain-language course with scrubbable animations: software, models, agents, bottlenecks, prompting, scope.

Choose a lesson
12 chapters

From zero to agent

The engineering book: runnable Python cells, provider adapters, memory, retrieval, MCP and production practice.

Open the book

What to expect, honestly

Check the source

Examples teach a method. Verify facts against original documentation before relying on them.

Expect change

Model capabilities, prices and APIs change. Treat version-specific examples as starting points, not guarantees.

It is not advice

The professional guides are education, not legal, medical or financial advice. You stay accountable.

It is hands-on

Explore visual explanations and a browser-based Python subset. Provider and tool integrations are teaching simulations.

Learn by experimenting

Inspect a simulated agent loop, change its inputs, and compare the output. No API key needed.

Back to the lab