Georgiy Lebedev is a systems-focused software engineer and PhD researcher specializing in database systems, with eight years of hands-on experience in low-level systems programming, compiler tech, and storage engines. He spent several years as a core contributor and backend developer for the widely used Tarantool DBMS, optimizing MVCC transaction handling, LSM/LSM-tree compaction, and wire-protocol behavior while mentoring student projects like HyperLogLog and buffer pool implementations. Now a Doctoral Assistant at EPFL, his research and engineering work bridges rigorous numerical-methods background from MIPT and HSE with practical performance tuning—he even implemented an LLVM-based JIT for Tarantool’s DQL during a GSoC internship. Colleagues can expect deep C/C++ expertise in cooperative multitasking environments, subtle instrumentation like backtrace collection, and a knack for turning theoretical analysis into production-quality fixes.
8 years of coding experience
4 years of employment as a software developer
Doctor of Philosophy - PhD, Database Systems, Doctor of Philosophy - PhD, Database Systems at EPFL
Bachelor of Science - BS, Applied Mathematics and Physics, 4.83 (excellent), Graduation with Honours, Bachelor of Science - BS, Applied Mathematics and Physics, 4.83 (excellent), Graduation with Honours at Moscow Institute of Physics and Technology (State University) (MIPT)
High School Diploma, Physics and Mathematics, 5.0, High School Diploma, Physics and Mathematics, 5.0 at Lyceum «Second School»
Get your data in RAM. Get compute close to data. Enjoy the performance.
Role in this project:
Backend Developer
Contributions:1522 reviews, 155 commits, 136 PRs in 1 year 5 months
Contributions summary:Georgiy primarily focused on enhancing the Tarantool database's core functionality through code modifications and performance optimizations. Their contributions involved reducing log verbosity for snapshotting and improving the `say_ratelimit` macro to manage and emit warnings about suppressed messages more efficiently. Additionally, the user addressed issues related to configuration updates and transaction handling, implementing fixes for log configuration updates and correcting handling of tuple field count overflows. Furthermore, the user contributed to performance by optimizing vinyl database's compaction and adding checks.
Get your data in RAM. Get compute close to data. Enjoy the performance.
Contributions:1329 pushes, 130 branches, 12 comments in 3 years 9 months
closeperformanceramcompute
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.