Stock Forecast
End-to-end machine learning pipeline for stock market forecasting, with automated data ingestion, model training, experiment tracking, and scheduled retraining.
Python Apache Airflow MLflow DVC FastAPI Docker
About the project
Stock Forecast is an end-to-end machine learning pipeline for stock market forecasting. The goal is to automate the entire lifecycle — from data ingestion to model retraining — with a reproducible and production-ready setup.
Stack used
The project is built around a modular MLOps architecture:
- Apache Airflow for orchestrating and scheduling the data and training pipelines
- DVC for data versioning and pipeline reproducibility
- MLflow for experiment tracking and model registry
- FastAPI for serving predictions via a REST API
- Docker for containerising all services with
docker-compose