Modular assignment paths
Each workshop is broken into discrete assignments that build on each other. Participants can see where they are in the sequence and understand what each task is preparing them to do next.
About Tourismre
Most machine learning education stays abstract. Learners leave courses knowing the theory but struggling to apply it to something as nuanced as financial decision-making. That gap between concept and practice is exactly what this platform was built to close.
Where models meet real financial decisions
Tourismre is a workshop-based learning platform designed for practitioners who want to work with ML in the context of personalized financial recommendations. Each programme is structured around hands-on assignments, iterative experimentation, and collaborative problem-solving - not passive video consumption.
Structured around step-by-step applied assignments
Covers recommendation systems, feature engineering, model evaluation
Accessible from any region with equal programme access
The pedagogical structure behind each workshop
Financial recommendation systems are not a single algorithm. They involve data pipelines, user behaviour modelling, evaluation under real-world constraints, and ethical considerations around bias. The curriculum reflects that complexity without overwhelming learners from the start.
Iterative feedback loops
Submissions are reviewed with specific, actionable notes. Learners revise and resubmit, which mirrors how real model development actually works - not a single pass but a cycle of adjustment.
Collaborative problem spaces
Certain exercises are designed for small groups. Participants compare model choices, debate evaluation metrics, and challenge each other's assumptions - which sharpens reasoning faster than solo study.
Real financial datasets used throughout - not synthetic toy examples
Python-based exercises with reproducible notebooks
Evaluation covers precision, recall, and business-relevant metrics
Bias auditing included as a required component, not an afterthought
Asynchronous-first design for participants across different time zones
Platform reach and programme depth
9
provinces served
14
workshop modules available
6
years of programme refinement
Participant background distribution
| Finance | 62% |
|---|---|
| Data roles | 51% |
| Engineering | 38% |
| Other fields | 28% |
Workshop completion by module type
| Solo | 81% |
|---|---|
| Group | 76% |
| Capstone | 68% |
The people and principles behind the work
Naledi Osei-Bonsu
Curriculum Lead
"I spent years watching capable people fail to apply ML because their training never touched real financial data. That is what we set out to fix - one assignment at a time."
Benedikt Mwangi-Zietsman
Technical Programme Director
"Recommendation systems in finance are deceptively hard. Getting the evaluation criteria right matters as much as the model architecture. We teach both, and we make sure learners understand why."
Every programme is fully asynchronous and bandwidth-considerate. Learners in rural provinces access the same materials and feedback as those in urban centres - no tiered experience.
ML applied to finance is genuinely complex. Programmes do not oversimplify. Learners are told upfront what background knowledge helps and where they are likely to struggle.
There are no lecture-only modules. Every concept is introduced through a task or experiment. If something cannot be demonstrated through practice, it is not included.
Financial ML is a moving field. Assignments are reviewed after each cohort and updated when tools, regulatory expectations, or evaluation standards shift meaningfully.
