Xahcajo

This project was inspired by the need to help sales teams understand why opportunities are won or lost. We combined a machine-learning model with explainability methods and a generative AI layer to turn complex insights into simple, human-readable explanations.

We learned how to manage real data, align different explainability techniques, and build a smooth end-to-end pipeline. Our main challenges were handling inconsistent outputs and keeping explanations both accurate and easy to read.

Built With

• fastapi
• gemini
• python
• streamlit