How r-suresh-kumar/udacity-devops-ml-api works
Project type and purpose
• This is a machine-learning microservice API written in Python (Flask) that serves pre-trained sklearn housing-price predictions.
• It is fully containerized with Docker, deployed to a local Kubernetes cluster, and continuously tested & built via CircleCI.
File structure
• Top-level files & dirs:
• app.py ← Flask application exposing a /predict endpoint
• model_data/ ← serialized model (.joblib) and source CSV
• Dockerfile, run_docker.sh, make_prediction.sh ← container build & run scripts
• run_kubernetes.sh, upload_docker.sh ← Kubernetes deploy & CI upload scripts
• .circleci/config.yml ← CI pipeline definition
• Makefile & requirements.txt ← build, lint, and dependency management
• No external DB—model loaded from local file, logs printed to stdout.
README for additional insights
• Confirms Docker + Kubernetes deployment steps
• CI goal: linting, build Docker image, push, deploy to k8s, test inference
• Emphasis on “operationalize production microservices.”