Presentation
My focus is high-throughput distributed systems and real-time streaming. The most concrete example: at Alo Application I rewrote the market-feed parser in unsafe C#, cutting parse time from 30 microseconds to 44 nanoseconds (700x faster). That removed the main bottleneck and let the system reach 500,000 Kafka messages per second in production.
I like the hard problems: race conditions, snapshot consistency, protocol parsing at the binary level, and cache invalidation that actually works under load. I use DDD, CQRS, and Clean Architecture when the problem warrants them, not as defaults.
Outside work I maintain ThunderPropagator, a multi-protocol .NET 10 streaming library published on NuGet. It supports WebSockets, MQTT5, QUIC, WebTransport, and TCP/UDP across a 6-repo monorepo with 970+ commits.
Open to senior .NET backend and full-stack roles, remote or hybrid.


