For full course details, download the comprehensive brochure ➡️

Course Outline

This is a build course. Each of the seven sessions opens with a short concept block and spends the rest of its three hours in the editor: you leave every session with an agent that runs, a test that proves it runs, and a note on what it costs. The arc goes from a single tool-using assistant to a multi-agent investment committee, then through interpretability, retrieval and reinforcement learning layers to a benchmarked prop-desk crew, and finally the evaluation and governance work that decides whether any of it is fit for a controlled pilot.

• Start Date: Monday 12th October
• Limited Availability: Enrol today – Applications close 30th September 2026.
• Delivery Mode: Live Online (flexible learning over 7 weeks, 1 lecture per week)
• Structured Assessment: Practical labs & Final Project
• Total Lecture Hours: 28 hours
• Instructors: Miquel Noguer i Alonso, PhD, ARPM — Founder & Chief Science Officer, AIFI & David Pacheco Aznar, Staq.io
• Discount:
  • Super early bird discount: 20% until 28^th August 2026
  • Early bird discount: 10% until 25^th September 2026

Labs use production frameworks — OpenAI Responses SDK, AutoGen, CrewAI, LlamaIndex, LangChain and Hermes — alongside the emerging RL–tool-use libraries (ARTIST, ReCall, verl) and simulators (TradingAgents, FinRL-Meta). Hermes — the open agent framework from Nous Research — is the reference harness: labs run through its Python interface in stateless-parser mode at every proposal boundary, meaning an empty toolset whitelist, no persistent memory and strict JSON output each time the agent proposes an action. The deterministic core, audit trail and risk gate stay framework-independent.

---

Tuition Includes

✓  21 core instructional hours across seven sessions
✓  Two faculty across the seven sessions
✓  Seven optional 60-minute office hours, Mondays after class
✓  Starter repository, notebooks, eval harness and a dated reading list
✓  Six lab systems plus your evaluated capstone
✓  AIFI Professional Certificate upon successful completion

---

Testimonials:

“My main objective was to get closer to finance topics. I come from the engineering world I studied telecommunications engineering and later moved into artificial intelligence engineering. But I was applying powerful algorithms like deep reinforcement learning in the telecommunications sector. I felt I needed financial background, and this course is not only a powerful strategic tool for AI, but it also provides excellent finance content.

One of the best aspects was the number of examples. There were many code snippets, demonstrations, and practical illustrations. The professors explained everything clearly, including how to adapt and optimise algorithms. That was one of the most satisfying outcomes of the course.”

“When people ask me which AI agents course I would recommend, I recommend this one because the faculty are genuinely helpful. There are multiple faculty members, and even though they are busy and highly sought after, someone always responds. You can also continue to access their help after the programme, which is really valuable. If you want assistance with projects later on, you can still reach out to them, and that’s something I really appreciated.”