Learning
AI, explained simply — one concept at a time.
I'm Josyula. I'm learning AI from scratch. I'll explain every concept here as simply as I can, as I learn it. No jargon walls, read in order, mistakes included.
The curriculum · 8 modules
✓
Module 01
ML Foundations Live
Industry-depth foundations: learning types, pipeline, metrics, splits, overfitting, and leakage — in plain English.
✓
Module 02
Classic Algorithms Live
Linear/logistic regression, trees, bagging vs boosting, baselines, and when tabular models beat deep nets.
✓
Module 03
Deep Learning Live
Neurons, activations, backprop, optimizers, regularisation, CNNs — and how this becomes the engine behind GenAI.
✓
Module 04
Transformers & LLMs Live
Tokens, self-attention, the transformer stack, and how models like ChatGPT are trained.
✓
Module 05
Generative AI Live
Prompting, RAG, tools, hallucinations, and how to evaluate GenAI products.
✓
Module 06
Data & Feature Engineering Live
Pipelines, leakage-safe splits, features, imbalance, and labels that decide model success.
✓
Module 07
MLOps & Deployment Live
Serving, versioning, rollouts, monitoring, drift — and LLMOps extras.
✓
Module 08
System Design & Interviews Live
A 6-box design framework, worked examples, and how to defend your choices in interviews.
Roadmap, projects & interview prep
The modules are the lessons. The roadmap is the plan. The projects are real Finance /
Energy systems you can demo in interviews.
Plan
From zero to interview-ready
Week-by-week phases, concepts, math, and checkpoints mapped to all 8 modules.
Build
Finance & Energy projects
Real-domain builds for interviews — fraud, filings RAG, load forecasting, and more.
Drill bank
Questions you’ll actually get asked
ML, deep learning, LLMs, MLOps, and system-design prompts with plain-English answers.
About this project
This is a learning log, not a textbook. I'm teaching each idea the way I wish it had
been taught to me on day one — plainly, with a concrete example, and an honest note
when something confused me. If you spot a mistake, tell me; corrections are the point.
Following along on X: @IamJosyula ·