The Optiver Trading App is designed to predict closing price movements of Nasdaq-listed stocks using order book data and various machine learning techniques. The project utilizes cloud engineering principles to ensure scalability, flexibility, and real-time processing, making it suitable for high-stakes environments where accurate price predictions are critical.
- Real-Time Data Processing: Utilizes AWS Kinesis Data Streams and Lambda for real-time data ingestion and processing.
- Scalability: Easily handles large datasets with the ability to scale resources up or down.
- Machine Learning: Implements XGBoost for incremental training and prediction.
- Flexible Cloud Architecture: Integrates various AWS services (ECS, RDS, S3) for a robust infrastructure.
- Web Interface: Provides a user-friendly interface for interacting with the application and visualizing performance.
The Optiver App is the core component responsible for processing real-time data and making predictions based on the trained machine learning models.
- For detailed documentation, refer to the Optiver App Documentation.
- For detailed API Reference, refer to the Optiver DB APIs Documentation.
The Train App handles the training of machine learning models using historical order book data. It ensures the models are up-to-date and accurate.
- For detailed documentation, refer to the Train App Documentation.
- For detailed API Reference, refer to the Train-Inference APIs Documentation
The Frontend provides a web interface for users to interact with the system, visualize data, and view predictions. For detailed documentation, refer to the Frontend Documentation.
The Data Streaming component is responsible for real-time data ingestion using AWS Kinesis Data Streams. It ensures that live market data is continuously fed into the system for processing and analysis. For detailed documentation, refer to the Data Streaming Documentation
- Ayush Agarwal
- Kevin Li
- Kexian Wu
- Mark Li