MLDeployLab
An end-to-end workflow demonstrating the building, deploying, and serving of machine learning models using FastAPI, Docker, and Streamlit.
The Narrative
Developed an end-to-end machine learning pipeline covering model training, API deployment, and interactive frontend serving.
Architected a FastAPI backend to expose RESTful endpoints for real-time model inference using PyTorch models and serialized encoders.
Built an interactive Streamlit frontend for users to easily test classifiers (like Animal and Rice Type Classifiers) and submit feedback.
Containerized both frontend and backend services using Docker and orchestrated them via Docker Compose for reproducible local and cloud deployments.
System Highlights
FastAPI Backend
High-performance REST API handling real-time inference requests and model serving.
Streamlit UI
Interactive web interface for seamless model testing and user feedback collection.
ML Models
Deployed trained PyTorch and scikit-learn models for image and tabular data classification.
Dockerized
Multi-service architecture fully containerized for consistent deployment across environments.
