πŸ“š Study Notes / Home / GenAI
Subject 01 Β· Generative AI (GenAI)

Generative AI (GenAI)

How tools like ChatGPT really work β€” tokens, transformers, prompting, agents, RAG, and deployment.

πŸ“„ 1-page cheat sheet πŸ“š 13 sessions ⏱ ~4.9 h total reading πŸ“„ Original course notes

πŸ“Ž Assignments & course material (original docs shared in the course)

SESSION 01

How AI like ChatGPT actually works

⏱ 25 min read Β· πŸ“– 8 topics

This is your very first GenAI class.

Read notes β†’
SESSION 02

How to actually talk to an AI so it does what you want

⏱ 25 min read Β· πŸ“– 9 topics

In Session 1 we learned how an LLM works on the inside.

Read notes β†’
SESSION 03

Giving the AI hands: from chatbot to agent

⏱ 23 min read Β· πŸ“– 7 topics

In Session 1 you learned that an LLM is just a next-token predictor that lives in a frozen brain with no memory and a knowledge cutoff.

Read notes β†’
SESSION 04

Giving an AI an open book: how RAG works

⏱ 23 min read Β· πŸ“– 7 topics

In Session 1 you learned that an LLM only knows what it absorbed during training β€” so it has a knowledge cutoff , can't see your private files, and…

Read notes β†’
SESSION 05

Making RAG actually good

⏱ 27 min read Β· πŸ“– 8 topics

Last time (Session 4) you built a basic RAG system: store documents, find the relevant ones, and feed them to the model.

Read notes β†’
SESSION 06

Building real apps: LangChain & LangGraph

⏱ 24 min read Β· πŸ“– 8 topics

So far you've learned what a model is (Session 1), how to talk to it well (Session 2), how to give it tools and turn it into an agent (Session 3),…

Read notes β†’
SESSION 07

Building agents the easy way: the Agent SDK

⏱ 22 min read Β· πŸ“– 7 topics

In Session 6 you wired up agents by hand with LangChain and LangGraph β€” drawing every node and edge yourself.

Read notes β†’
SESSION 08

Threads & Tracing β€” keeping memory of, and watching, your agent

⏱ 22 min read Β· πŸ“– 5 topics

In Session 7 you built your first agent with the Agent SDK.

Read notes β†’
SESSION 09

How AI agents remember things

⏱ 20 min read Β· πŸ“– 5 topics

Back in Session 1 we learned a surprising fact: the model itself has no memory .

Read notes β†’
SESSION 10

Teaching AI to remember connections, not just facts

⏱ 17 min read Β· πŸ“– 4 topics

In Session 9 we gave our agent a memory; in Sessions 4–5 we taught it to search documents with vectors.

Read notes β†’
SESSION 11

Talking to AI out loud β€” how voice agents work

⏱ 20 min read Β· πŸ“– 5 topics

You've already met the AI "brain" (the LLM from Session 1) and seen how it can act as an agent that uses tools (Sessions 3 & 7).

Read notes β†’
SESSION 12

How AI agents plug into tools and talk to each other

⏱ 24 min read Β· πŸ“– 8 topics

In earlier sessions you taught a single agent to call tools and hand work to other agents.

Read notes β†’
SESSION 13

Shipping your GenAI app to the real world

⏱ 23 min read Β· πŸ“– 7 topics

This is the final class.

Read notes β†’