Zero-Downtime Trading Platform Infrastructure
2023 – 2026
Solo-built cloud infrastructure for a B2B energy trading platform, designed for zero downtime from day one.
Impact
- ›Auto-scaling infrastructure across multiple availability zones with automated health checks and self-healing recovery
- ›Real-time ingestion of high-frequency multi-sensor telemetry alongside EU energy market data feeds
- ›Infrastructure-as-code from day one, enabling fast, repeatable recovery and deployment
As founding engineer, I designed and built the infrastructure layer for a platform handling live financial transactions in energy trading — where downtime directly translates to missed trades and lost revenue.
The problem
Commercial and industrial battery operators needed a platform that could ingest live sensor telemetry and market data continuously, run automated trading decisions on top of it, and stay up. As a solo founding engineer building this from scratch, the infrastructure had to be resilient without a dedicated ops team to babysit it.
What I built
I designed the system around auto-scaling EC2 instances spread across multiple availability zones, sitting behind a load balancer with continuous automated health checks and self-healing recovery. All of it was defined as infrastructure-as-code in Terraform, so the entire environment could be rebuilt or recovered quickly and consistently rather than depending on manual intervention.
On the data side, I built a real-time ingestion pipeline combining OPC UA and MQTT sensor telemetry with Kinesis streaming, alongside live market data feeds from ENTSO-E (the European electricity market data platform).