Gen & Agentic AI
The Gen AI course, charted end to end
18 lectures charted in the Captain’s Log, 13 with material up · 10 booklets, 432 pages, free to read
Captain Cortex
Keeper of the Neural Isle
Arr! Ye've found the Neural Isle, the one island charted from shore to shore. Machine learnin', neural nets, transformers, RAG an' agents, every page cross-checked against 20 hours o' class recordings. An' it costs ye nothin' -- sign in an' read the lot.
Below be 10 files, sorted by the day they were plundered:
Topics covered
- Linear regression, MSE, the normal equation
- Gradient descent and the learning rate
- Decision trees, Gini, Random Forest, boosting
- Neural networks and backpropagation
- Embeddings, attention and transformers
- Temperature, model choice and cost
- RAG, chunking, retrieval, grounding
- Prompt engineering
- Fine-tuning, LoRA and QLoRA
- Agents and the ReAct loop
The Whole Voyage
Start here if ye have one night left. Sit the mock paper first.
Revision: the whole workshop
All four days in one booklet, written for someone starting from nothing. 15 sections, every formula explained with a worked example, every correction he made, and a revision plan. Colour-coded: red is must-know, orange is very likely in the exam. The Practice Booklet is included free.
66 pages · 3.8 MBFree₹149Practice: 113 questions, answers at the back
A full 50-mark mock paper plus three question banks. Questions first, every solution worked line by line at the end, so you can sit it like the real thing. Free with the Revision Booklet.
18 pages · 1.4 MBFree₹49Companion: what is NOT in the exam
The capstone brief and rubric, Dify, the business framing, and setup, lifted out of the revision booklet so that booklet is only examinable material.
14 pages · 0.2 MBFreeDay 1, Machine Learning from nothing
Regression, trees, forests, boosting.
Day 1 handwritten notes
ML from nothing: linear regression, the normal equation, gradient descent, overfitting, trees, Random Forest, boosting. Every number traced by hand.
47 pages · 14.9 MBFree₹99ML deck companion
All 19 slides of his ML deck, with the figures and the errors found.
58 pages · 3.3 MBFree₹79Day 2, Python, neural nets, embeddings
The full Python morning, then the maths.
Day 3, Transformers, cost and RAG
Backprop corrected, then retrieval.
Day 4, Prompting, fine-tuning, agents
RAG as code, LoRA, and the capstone.