Skip to content

Generative AI for Managers

Generative AI for Managers - TUHH Institute of Entrepreneurship · part of my Technology Management MBA · study notes for revision.


This was an intensive 6-day, hands-on course on Generative AI, built for business students with no or low coding background - exactly the position I was in. It goes from how LLMs actually work all the way to building a GenAI business case, with a lot of practice in between using Google Colab and the Gemini API.

These are study notes for revision: the concepts explained plainly, the jargon defined the first time it shows up, and the hands-on labs summarised with the notebooks attached for download so I (or anyone) can open them in Colab and actually run them.

Setup - Colab & the Gemini API

Get a working environment in ~15 minutes: Google Colab, a Gemini API key stored safely in Colab Secrets, the google-genai SDK, and a first test call - plus a troubleshooting table.

Read notes →

Day 1 · LLMs & Transformers

What large language models really are - next-token prediction, tokens and embeddings, the transformer and attention, how models are trained (pre-training → instruction tuning → RLHF), and the practical knobs: tokens, context window, temperature. Ends with capabilities vs. limitations.

Read notes →

Day 2 · Prompting & Extraction

Systematic prompting techniques for reliable results, and pulling structured data out of messy unstructured text.

Read notes →

Day 3 · Retrieval-Augmented Generation

Grounding an LLM in your own documents - building a knowledge base with embeddings and a RAG pipeline for accurate, source-backed enterprise Q&A.

Read notes →

Day 4 · GenAI Agents

LLMs that take actions - designing autonomous agents with tool use and multi-step reasoning.

Read notes →

Day 5 · GenAI-Assisted ML

Using GenAI and AutoML on business data, then interpreting and communicating the results to non-technical stakeholders.

Read notes →

Day 6 · GenAI Business Cases

Frameworks to spot where GenAI actually creates value, weighing benefits against cost and risk, and turning an idea into a GenAI product proposal.

Read notes →