FREE · HANDS-ON · FRONTIER AI
Learn frontier AI.
Build with it.
Know its limits.
Free, hands-on courses from how transformers work to RAG, evals, agents and MCP. Read the idea, run the code in your browser, and build projects you can show.
READ IT. THINK IT. TRY IT.
Your first data insight.
Three days. Three numbers. What is the average?
minutes = [12, 18, 30]
average = sum(minutes) / len(minutes)
print(average)Watch · 1 minute
Kishor Academy in one minute
What you learn, how the lessons work, and where to start.

Choose where to start.
Four doors in. Every course ends in projects you keep.
From first principles to agents.
One sequence from understanding how models work to shipping tool-using systems. Each stage links to a course when it is open.
- 01Open now
Introduction to LLMs
Begin with an available introduction to language models.
- 02Open now
Python & Data for Machine Learning
Prepare with Python, data, and practical programming.
- 03Open now
Large Language Models: From Prompts to Practice
Learn how to prompt and evaluate language models.
- 04Open now
LLM Engineering: RAG, Evals & Multimodal Apps
Design retrieval, evaluation, and an evidence-based application.
- 05Open now
Agentic AI: Tools, Workflows & MCP
Add bounded tools, approval steps, and failure recovery.
Read it. Run it. Check it.
Every lesson is a page you read, code you run, and a question that tells you whether it landed.
- 01
Python in the browser
Multi-file projects, input and arguments, run in an isolated sandbox. Nothing to install.
- 02
Reviewed projects
Courses include a portfolio project, scored against a published rubric by a person, not a bot.
- 03
Lesson-grounded help
Where enabled, ask about the lesson you are on. Answers link back to the lesson text and never grade you.
- 04
Certificates
Finish a course to get a printable certificate with a shareable verification page.
TERM OF THE DAY
MCP
Model Context Protocol: an open standard for connecting AI applications to tools and data through servers.
Browse the glossaryThe frontier, mapped.
Reasoning models, agents, post-training, evals, multimodal and safety, each explained with what to learn, which course covers it and the papers behind it.
Every lesson in English or বাংলা.
Switch language from the top bar. Code, terms and sources stay the same, so you can move between them.
এআই কী করতে পারে, কোথায় এর সীমাবদ্ধতা, এবং কীভাবে চিন্তাশীলভাবে ব্যবহার করবেন তা জানুন।Fair questions, straight answers.
Answers to the practical questions, so you can focus on learning.
Read the full guideHow do I start learning?
Browse the course catalogue, open a course, and review its curriculum, requirements, price and access period. Create an account or sign in to enrol. Your enrolled courses appear in My learning.
Can I write and run code here?
Lessons with reviewed interactive material include a Python code workspace. Open Code workspace, edit a file and select Run Python. Your draft stays in this browser; export your project to keep a separate copy. The workspace does not provide a full system terminal.
Are all courses free?
Free courses are labelled Free. This preview supports enrolment in published free courses with available lessons. Paid purchases are not enabled here yet.
Will every course include a certificate?
Certificate availability depends on the course and its completion requirements. Certificates are not issued by this preview. A completion certificate does not imply accreditation or a professional qualification.
How long can I access a course?
Access depends on the course and your enrolment. Check the access period on the course page and any expiry date in your account. Access is not automatically lifetime for every course.
How do lesson questions and practice work?
Read the lesson, answer any practice question, and use the discussion to ask questions or share a reflection. Practice feedback appears after you submit an answer. Mark each lesson complete when you finish.