Hi HN,
I built a small open-source web app that generates confusion matrices for ML models.
I originally needed this while working on a project and was surprised that many simple tools were either paid, cluttered, or required heavy setup. I wanted something lightweight, free, and easy to use.
The app supports numeric and string labels, CSV upload or manual entry, multi-class matrices, and shows common metrics like accuracy, precision, recall, and F1-score. You can also download the generated matrix as an image.
It’s fully open source and built with Streamlit. I’d love feedback, suggestions, or contributions from the community.
GitHub: https://github.com/pareshrnayak/confusion-matrix-generator
Live demo: https://confusion-matrix-gen.streamlit.app
Technical question: did you consider support for class imbalance scenarios (e.g. thresholding or cost-weighted confusion matrices)? In many real datasets, F1/accuracy hide useful signal unless you vary the decision rule.
Clean implementation, and good call on keeping it simple.