Shourya Katiyar

I’m a Software Engineer building cloud-native, low-latency systems for real-time trading. At JP Morgan Chase & Co., I architected a Kafka-, Deephaven-, and gRPC-based pipeline processing up to 250K trades per 10 seconds with sub-100ms latency.

I engineered a horizontally scalable ingestion layer on Rust running in AWS ECS Fargate, normalizing 10+ heterogeneous venue schemas while sustaining 5× traffic spikes without consumer lag buildup. I also optimized stateful position calculations using Deephaven incremental tables (Append → Trade → Position), cutting computation latency by 45% under peak load.

I design reliability and performance end-to-end: a secure, low-latency Java gRPC service with sub-50ms latency and 500 concurrent streams, plus an active-active multi-region architecture delivering 99.99% availability. I further reduced network overhead by 60% by replacing REST/SSE polling with gRPC streaming, and provision distributed infrastructure with Terraform to improve deployment consistency.