Real-Time Algorithmic Energy Trading Platform
I architected and built a high-performance, serverless data platform on GCP to transition a manual, batch-based energy trading operation into a fully automated, real-time algorithmic system. Using Python, Apache Beam, and Dataflow, I designed streaming pipelines capable of processing over 10 million daily events from diverse market, weather, and grid signals with sub-second latency. The entire ecosystem was containerized via Docker and managed through Terraform to ensure zero-ops scalability and perfect environment parity.
The transformation was game-changing for the client: we reduced data-to-decision windows from hours to under one second, enabling 24/7 automated trading while slashing infrastructure costs by 40% compared to their previous VM-based setup. Beyond the technical gains, the platform delivered 99.9% uptime and a comprehensive audit trail for regulators, providing a scalable foundation for the client to expand into new energy markets.
"Julius led the development of our cloud-native data platform for real-time algorithmic trading in energy markets. Starting from scratch, he and his team launched a production-ready MVP in under a year. His ability to understand complex requirements, translate them into business terms, and deliver effective solutions makes it a pleasure to collaborate with him. I strongly recommend working with Julius."
I was responsible for this project as part of my role as "Head of Machine Learning & GenAI - Google Cloud" at adesso SE.