We started because good finance education shouldn't depend on where you live
Since 2022, we've been building courses that make predictive modeling accessible to anyone with an internet connection and the drive to learn.
What we actually do
We teach predictive modeling for finance. Not the theoretical kind that looks good on paper but falls apart when you try to use it. We focus on methods that work with real market data, handle messy inputs, and produce results you can actually interpret.
Our courses walk you through building models from scratch. You'll work with actual datasets, deal with missing values and outliers, test different approaches, and learn to spot when a model is overfitting or when your assumptions don't match reality. We cover time series forecasting, risk assessment, portfolio optimization, and credit scoring.
The instructors have worked in quantitative finance and know what skills matter when you're building something that needs to perform consistently. They explain the math where it's necessary and skip it where intuition is enough. Every course includes exercises with real data and detailed feedback on your approach.
We started in South Africa but serve students everywhere. The platform is completely online, so your location doesn't matter. What matters is whether you're willing to put in the work to understand how these methods actually function.
How we build courses
Real implementation work
You'll write code that processes data, trains models, and generates predictions. Every assignment requires you to make decisions about preprocessing, feature engineering, and validation strategies that affect your results.
Actual market datasets
We use historical price data, economic indicators, company financials, and alternative data sources. You'll learn to handle the irregularities and gaps that come with real-world information instead of clean academic examples.
Performance measurement
You'll track accuracy metrics, analyze error distributions, run backtests, and learn to distinguish between models that look good in training and ones that generalize properly to new data.
Method comparison
We cover linear regression, ARIMA, GARCH, random forests, gradient boosting, and neural networks. You'll understand when each approach makes sense and when simpler methods outperform complex ones.
Risk consideration
Every model has limitations. You'll learn to identify assumptions, test robustness, estimate uncertainty, and communicate what your predictions can and cannot tell you about future outcomes.
Iterative refinement
Building good models takes multiple attempts. You'll submit work, receive specific feedback on your methodology, adjust your approach, and resubmit until your implementation meets professional standards.
What drives our work
Education shouldn't have geographic barriers
We believe quality finance education should be available to anyone with internet access, regardless of where they were born or where they currently live. Our platform works the same whether you're in Johannesburg, Mumbai, or São Paulo.
We keep enrollment open year-round so you can start when it fits your schedule. Course materials remain accessible after completion so you can reference them when working on real projects. No artificial scarcity, no pressure tactics.
- All lectures include transcripts and downloadable code examples
- Forums stay active with peer discussion and instructor responses
- Pricing remains consistent across regions with purchasing power adjustments available
- No prerequisites beyond basic programming and statistics knowledge
Theory serves practice, not the other way around
We explain the math when it helps you understand how a method works or why it might fail. We skip mathematical proofs that don't change how you'd implement or interpret the model. The goal is competence, not academic completeness.
Every concept connects to a specific application. You'll learn stationarity testing because non-stationary data breaks ARIMA models. You'll study regularization because overfitting is a real problem when you have more features than samples.
- Assignments use datasets from actual markets and companies
- Projects require end-to-end implementation from data loading to prediction
- Evaluation criteria focus on model performance and code quality
- Case studies examine both successful predictions and notable failures
We don't oversell what modeling can do
Predictive models are tools with limitations. They work better for some problems than others. They require careful validation and ongoing monitoring. They can't eliminate uncertainty or guarantee profits.
Our instructors discuss when models failed in their own work, what went wrong, and how they adjusted. You'll learn to recognize overconfident predictions, identify when you're data mining rather than discovering patterns, and communicate uncertainty honestly.
- We cover common pitfalls like look-ahead bias and survivor bias explicitly
- Courses include examples of models that performed well in testing but poorly in production
- Assignments require documenting assumptions and limitations alongside results
- No promises about returns, success rates, or job placement guarantees
Courses evolve based on what students need
We update material when new methods prove useful or when feedback reveals confusing explanations. Student questions often highlight gaps in coverage that we address in revised versions.
Instructors monitor forums to identify common struggles. If multiple students misunderstand a concept, we record additional explanations or add exercises that build intuition more effectively.
- Course content updates reflect current best practices in quantitative finance
- We add new datasets when markets change or new data sources become available
- Exercises get refined based on which ones effectively build understanding
- Feedback from working professionals helps keep material relevant