AI & Machine Learning

Understand the landscape from scratch — what AI, ML, and deep learning actually are, how they differ, supervised/unsupervised/reinforcement learning, and how every major company uses them.

The minimal math that makes ML stop being magic: vectors as data, dot products as similarity, gradients as direction, and probability as honesty.

Regression, trees and clustering — the algorithms that still win on tabular data, plus the evaluation discipline (precision, recall, overfitting) that carries to all of ML.

From single perceptrons to deep networks — how neural networks learn, backpropagation, activation functions, CNNs for images, and building your first network in Python.

How LLMs work — from tokens to transformers to in-context learning. What prompting, temperature, and context windows mean, and how to build production applications with them.

Give LLMs access to your own data — how RAG works, vector embeddings, chunking strategies, re-ranking, and building production RAG pipelines.

Autonomous AI systems that plan, use tools, and execute multi-step tasks — how agents work, the ReAct pattern, tool use, memory, and building your first agent.

How vector databases store and search embeddings — HNSW, IVF, product quantization, and a comparison of Pinecone, Weaviate, ChromaDB, and pgvector.

Teaching a pre-trained model new behavior — full fine-tuning vs LoRA, when to fine-tune vs RAG vs prompting, data preparation, and evaluation.