What the programme covers
Each client works through a structured sequence of modules, adjusted to their existing knowledge and professional context. The pace and depth are set in the first session.
Sessions are conducted remotely on a fixed weekly schedule. Between sessions, clients work through practical assignments reviewed before the next meeting.
Reading financial signals with machine learning
- Supervised models for time-series financial data
- Feature engineering from raw market and accounting inputs
- Backtesting logic and avoiding look-ahead bias
Forecasting and uncertainty quantification
- Probabilistic output models - intervals over point estimates
- Gradient boosting and ensemble approaches in practice
- Handling missing data, outliers, and regime shifts
Embedding models into existing workflows
- Audit of the client's current data infrastructure
- Selecting the right deployment pattern for the context
- Monitoring model drift and retraining triggers
- Documentation standards for non-technical stakeholders
Format and delivery
All sessions take place over video call. Clients receive recorded replays, written session notes, and access to a private resource library between meetings.
The programme is available to clients worldwide. No prior machine learning experience is required - the entry point is calibrated individually during a free scoping call.
Scheduling is flexible across time zones. Most clients in professional roles complete the programme alongside full-time work without significant schedule disruption.
Clients on the programme
Three practitioners describe what working through the programme looked like in practice. Results vary by starting point and application context.
Oisín Farragher
Portfolio Risk Analyst"I came in knowing Excel well and Python poorly. By week six I had a working forecasting pipeline on real portfolio data. The pace was demanding but the assignments were directly relevant to my job."
Dagmara Wróbel
FP&A Lead"The integration module was the most useful part for me. I didn't need to build models from scratch - I needed to understand how to work with the ones my team was already producing. That's exactly what we covered."
Rúna Sigurðardóttir
Independent Consultant"Working across time zones from Reykjavík was never an issue. Sessions were recorded, notes arrived by the next morning, and the specialist adjusted the syllabus twice when my client project shifted direction."