Hi! I'm Vladislav Oleshko.

Systems engineer specializing in high-performance C++, databases, storage, concurrency and I/O.

Fascinated with programming and computers since my childhood, winning competitions in my teenage years and building new technologies in my career.

📍 Warsaw, Poland

GitHub LinkedIn

C++ Rust Go Python     Concurrency Fibers & coroutines Async I/O io_uring Linux   Performance engineering Profiling Benchmarking Memory management Data structures   LLM agents / MCP

Core Software Engineer at DragonflyDB August 2022 - present (4 years)
Multithreaded asynchronous Redis replacement written in C++
First engineering hire — helped take Dragonfly from early prototype to a trusted in-memory database with hundreds of thousands of deployments worldwide
Development
• Designed and implemented core components from scratch: replication, structured and vector search, data tiering
• Led SSD data tiering from prototype to general availability: io_uring-based offloading, packing small values into shared pages, defragmentation, snapshotting, replication and clustering
• Optimized and extended multithreaded transaction runtime supporting millions of ordered (serializable) operations per second
• Extended helio , the open-source fiber runtime behind Dragonfly: scheduler, synchronization primitives, io_uring abstraction
• Built randomized data generation and consistency-checking test infrastructure for replication and snapshots
• Reviewed code and shaped designs across core areas with the team, planned refactorings, drove adoption of new C++ features (like coroutines)
Systems & Optimization
• Profiled and optimized microsecond-scale hot paths (caching, prefetching, reordering, contention, allocations, etc)
• Optimized I/O patterns: provided buffers, scatter-gather and batched submissions, fewer copies and syscalls
• Ran multi-node latency benchmarks that correct for coordinated omission, covering scenarios like mixed workloads under replication failover
Production
• Provided support and technical guidance to customers, rolled out new features to terabyte-sized clusters
• Investigated incidents such as data discrepancies under replication failover, crashes, OOMs and deadlocks
Agents & MCP
• Developed MCP server for safe access to production observability and control plane data, optimized friendliness for LLMs (context savings, grouping, summaries)
• Created agents for problem root cause recognition and automatic incident investigation
Company
• Contributed to quarterly roadmap planning: scoping and prioritizing core features and projects
• Wrote technical blog posts on transactions, search, SSD data tiering and Lua scripting

Software Engineering Intern at Yandex May - September 2021
Yandex.Maps team responsible for public transport navigation
• Implemented C++ MapReduce algorithms for distance calculations running on millions of objects
• Implemented algorithms for bus route change detection and route estimation


Education: Belarusian State University, Minsk, Bachelor of Science in Informatics and Applied Mathematics   2020-2024

Awards:
• Belarus National Olympiad in Informatics, 2nd-degree diploma (2019, 2020)
• Russia Open High School Team Programming Contest (VKOSHP), 3rd-degree diploma (2018, 2019)
• Regional Intel ISEF qualification award for a VR app (2018)

Occasional open source contributions (Before AI!) 2022
• Made my own two Go small libraries: Redis streams-to-channels and a universal struct mapper
• Wrote a small article about ray-tracing with Rust
• Hand optimized x86 compiler builtins (memcpy and memset) for wazero - a wasm runtime
• Added a cache size limiter with automatic eviction to Fiber
• Implemented asynchronous shader compilation for Magnum - a C++ graphics library
• Added optimistic locking and embedded structs to Rel - a Go ORM

Other:
Proficient in English, German (lived there)