This project builds an income classifier based on the publicly available Census Bureau Dataset. The project incorporates automation through CI/CD. Furthermore, the produced model is checked for bias using Slice Testing. Finally the model is deployed to cloud using FastAPI on Render hosting service.
The project is setup with Continues Integration (CI) using Github Actions to run pytest for unit testing and API testing, and flake8 for linting. Furthermore, Continues Deployment (CD) is configured on the cloud to automatically deploy the application after every push to the repository provided the CI passes.
See the Model Card for details about the model training process, performance, biases, and caveats.