I’m a backend engineer working on high-performance and distributed systems.

Recently, I redesigned a time-series processing component from Python to Rust, reducing execution time from hours to minutes.

I tend to approach performance as a design problem, not a scaling problem — focusing on data flow, execution models, and technology choices rather than adding more infrastructure.

Currently, I’m building a distributed platform for real-time processing, simulation, and analysis of time-series data. The system is designed as a set of independent services with an emphasis on determinism, predictable performance, and extensibility.

In this work, I focus on:
– designing modular service architectures for high-frequency data
– building in-memory processing pipelines for simulation and analytics
– improving performance by rethinking execution paths
– optimizing latency, throughput, and resource usage under load

Recently, I’ve also been exploring LLM-driven workflows, where generated outputs are treated as intermediate representations and stabilized through deterministic backend systems.