At Mudrex, we use the best in class tech which helps us build fault-tolerant, highly scalable and concurrent systems which to scale to 10000's of strategies running at a tick level in a cost-effective manner.
Our tech stack uses:
- Python, Java, and Node for backend
- Core DB is in MySQL
- Time series data for portfolio analysis etc. is stored in Mongo using a compression algo called Arctic on top
- Queuing infrastructure is built using Kafka, Celery and Rabbit MQ
- We use Redis for a universal cashing layer to deliver cached responses quickly
- Socket.io for web sockets to communicate and transfer data to front-end
- ELK for alerting and monitoring
- Kong for handling access controls for our API platform
- React is being used for our web front end
- We are deployed on AWS and use a combination of EC2, S3, and Lambda to process requests
This ensures that we are able to process requests and place orders in the fastest way possible.