Project detail
Training Impact with PyMC
Bayesian evaluation of the effect of training programs on skills before and after.
Highlights
- Before/after training comparison
- Employees with significant improvement
- Results visualization and per-record detail
- Useful approach for L&D
No public repository or demo link is published for this project yet.
Architecture
Bayesian workflow with a FastAPI backend and Flutter app to evaluate training impact on employee records
Challenges
- Measuring real improvement rather than noise
- Presenting results in a comprehensible way
- Separating meaningful cases from minor changes
Outcome
A more solid way to justify training programs with evidence
Overview
Problem
Training is often evaluated without enough rigor, making it hard to distinguish real improvement from normal variation.
Approach
I compared skills before and after using a Bayesian model to estimate impact.
Outcome
The project helps measure which programs truly improve capabilities.