Your Model Isn't Done Until Someone Else Can Call It

A locally functional churn prediction model built with FastAPI is made accessible to external users through deployment. The process involves containerizing the application using Docker to create a self-sufficient image, which resolves environment and dependency issues. This Docker container is then deployed on an AWS EC2 instance, detailing the server setup, security group configuration, and troubleshooting connection problems. The container is run in detached mode and assigned a stable Elastic IP, making the model's API endpoint publicly callable. While functional, the deployment still lacks production-grade features like HTTPS and authentication.