Building high-performance systems at a quantitative investment firm. Focus on large compute, distributed systems, performance, and resiliency.

Member of File Systems and Networking Team, focused on improving Windows Server performance for networking workloads. Removed major throughput bottlenecks on large NUMA servers, addressing false sharing and lock contention. Created scalable reference and performance counter libraries, improved kernel locks and synchronization objects, and drove workflow design improvements for better batching and reduced latency.

Design, research, and development of HydraStor, a failure-resilient distributed system with asynchronous consensus procedures and local load optimization in heterogeneous architecture.

Designed and developed a monitoring dashboard HTTP server in Python with background jobs for data gathering and visualization. Contributed to garbage collector design for large Bigtable instances, a performance-critical C++ project.

Research on trade-offs between approximation factor and time complexity for NP-hard problems.

Topics: Maximum Independent Set, Vertex Coloring, Set Cover, and Bandwidth optimization.

Polish: Native or bilingual proficiency
English: Full professional proficiency