Knowledge Base
Reusable references for the whole course
This is the course’s knowledge base: standalone pages you can come back to during any week (or after the course). Weekly lessons link into this section; the material here is meant to outlive any single class.
AI & LLM fundamentals
- LLM Glossary — technical terms used in lectures and class.
- Which AI to use — January 2026 comparison of major models and tools.
- Beyond the course — curated videos, papers and further reading. Updated regularly.
Setting up AI tools
- Terminal Basics — essential shell commands for CLI work.
- Installing AI CLI Tools — Claude Code, Gemini CLI, Codex CLI, OpenCode.
- VS Code + Copilot setup — IDE + AI assistant.
Working with APIs
- How to get AI API keys — OpenAI and Anthropic.
- Introduction to APIs — fundamental concepts: what a call looks like, headers, auth, rate limits.
- Calling LLM APIs from Python — practical first calls, helper functions, and
pandasworkflows. - How APIs work under the hood — idempotency, retries, batching, tokens.
- Walkthrough: World Bank + FRED — gentle intro using public economic data APIs.
- Walkthrough: FBref football data — slightly harder end-to-end example.
Doing the analysis
- Designing larger analytics projects with AI — mindsets, self-interview,
agents.md, SKILLS, the three kinds of tests. - Documentation and READMEs — what good documentation looks like.
- Joining data tables — keys, joins, pitfalls.
- Creating graphs — iterative refinement from basic to publication quality (R example).
- PDF guide — working with PDFs.
- Good report ideas — what separates a solid report from a weak one.
Case studies
Datasets and code organised by case study — used across multiple weeks.
- World Values Survey — Weeks 2–3 (documentation, reports).
- Austrian Hotels — Week 4 (joining tables).
- Football Manager Interviews — Weeks 7–8 (text as data).
- US Earnings (CPS) — Week 6 (from data to report).
- Employee Commits — data-analysis demos.
- Common Support (R) — iterative AI coding.