Artificial Intelligence for Business and FinTech
Building, testing and governing AI systems for data, decisions and markets
ESE Short Course · Term 1, A.Y. 2026–2027 European School of Economics, Florence · one-to-one tutorial and lab · 30 contact hours
Niccolò Salvini, PhD, Adjunct Professor
Everything for this course is on this page. One row per session: the Colab notebook we build in class, the concept cards shown on the projector, and the homework that opens the following Wednesday. Materials are published by Thursday; homework is due Tuesday 23:59 in your Drive folder.
Sessions Wednesdays 10:00–13:00 · ESE Florence · notebooks open in Google Colab
| # | Date | Theme | Materials | Homework |
|---|---|---|---|---|
| 1 | Wed 23 Sep | How we will work · from one neuron to an LLM · reading a model release | slides · Colab ↗ · notebook | HW1 · due Tue 29 Sep |
| 2 | Wed 30 Sep | Inside an LLM — tokens, embeddings, attention, the transformer, temperature · a real model in Colab | slides · Colab ↗ · notebook | HW2 · due Tue 6 Oct |
| 3 | Wed 7 Oct | Look first, then predict · exploration, a black box, a decision | slides · anatomy of a SaaS · Colab ↗ · notebook | HW3 · due Tue 13 Oct |
| 4 | Wed 14 Oct | LLM applications I — API, structured output, RAG · EU AI Act | slides · Colab ↗ · notebook | HW4 · due Tue 20 Oct |
| 5 | Wed 21 Oct | LLM applications II — agents, tool use, evals · the product lens | slides · Colab ↗ · notebook | Case brief (40%) · due Sun 1 Nov |
| — | 26 Oct – 1 Nov | Reading week — no class · minor applied case brief | — | brief |
| 6 | Wed 4 Nov | Minor presentation · fintech landscape · capstone scoping | slides · Colab ↗ · notebook | HW6 · due Tue 10 Nov |
| 7 | Wed 11 Nov | Markets and backtesting done honestly · model risk · date may move | slides · Colab ↗ · notebook | HW7 · due Tue 17 Nov |
| 8 | Wed 18 Nov | Digital assets, on-chain data, DeFi · MiCA | slides · Colab ↗ · notebook | HW8 · due Tue 24 Nov |
| 9 | Wed 25 Nov | Responsible AI applied to the capstone · red-team | slides · Colab ↗ · notebook | HW9 · due Tue 1 Dec |
| 10 | Wed 2 Dec | Implementation and capstone studio | slides · Colab ↗ · notebook | Portfolio (60%) · due Sun 6 Dec |
| Rev. | 7–13 Dec | Final presentation — proposed Wed 9 Dec, 10:00 | — | portfolio brief |
Session 7, Wednesday 11 November. The tutor may be away that day for a commitment in Milan. If so, a make-up date is agreed in class and announced here and on Moodle.
Opening a notebook. Colab ↗ opens the session notebook in Google Colab. Then File ▸ Save a copy in Drive, into your ese-ai-fintech folder. The notebook link is a read-only preview of the same file.
How the course works the contract, stated once
- This is a lab. You write code from session 1, with an AI assistant at your side. You do not need to know Python; you need to be able to judge what the assistant produces.
- The rule of the course: anything an AI produced and you submit, you must be able to explain line by line. I will ask.
- Weekly rhythm: Wednesday 10:00–13:00 in Florence. The first 20 minutes are your walkthrough of last week’s homework. Notes and homework are published by Thursday; homework is due Tuesday 23:59.
- Assessment: a minor applied case brief (40%, after session 5, presented in session 6) and a final AI strategy and prototype portfolio (60%, revision week). Both are assessed on your notebooks and memos in your course folder. Details in the syllabus.
- Tools: Google Colab and Google Drive, one chat assistant of your choice, a Google AI Studio key from week 4. Nothing is installed locally. See Setup.
- Nothing confidential goes into an AI tool. All market, crypto and DeFi activities are educational simulations, not investment advice.
Tutor
Niccolò Salvini, PhD
Adjunct Professor