Roman Leventov is a founder and seasoned engineer with 14 years building low-latency, high-performance systems spanning concurrency, distributed data stores, and systems architecture. He blends deep algorithmic and JVM-performance expertise—evident from contributions to high-profile open-source projects like Chronicle, Chronicle-Map, and Apache Druid—with hands-on work optimizing serialization, memory, and hashing for real-world scale. Recently he’s focused on building personal, secure AI platforms and agent/knowledge infrastructure, applying the same systems-first approach to grounded inference and human/AI collaboration. Roman’s background includes running large Druid fleets, improving core database indexing and memory behavior, and shipping production LLM tooling; he also writes about technical and AI-safety topics on his engineering blog. An uncommon strength is translating low-level performance wins into higher-level product reliability, making him effective both as a contributor to critical OSS and as a founder.
14 years of coding experience
10 years of employment as a software developer
Bachelor's degree Computer Science, Bachelor's degree Computer Science at Bauman Moscow State Technical University
Replicate your Key Value Store across your network, with consistency, persistance and performance.
Role in this project:
Back-end Developer
Contributions:16 releases, 796 commits, 14 PRs in 4 years 1 month
Contributions summary:Roman's commits primarily focus on improving the Chronicle Map's replication API, including API cleanups and bug fixes related to replication and bootstrapping, and modifications to improve efficiency. The user also worked on ensuring that entries are correctly persisted and transferred. The changes involved creating new methods, improving the state management, and integrating code with the existing API.
Contributions:11 releases, 73 commits, 8 PRs in 5 years 2 months
Contributions summary:Roman primarily focused on enhancing the hashing functionalities within the Java project. They addressed minor Javadoc corrections, refactored an interface to an abstract class, and optimized hashing algorithms. They also added a MurmurHash3 implementation and integrated boolean hashing, demonstrating their involvement in core algorithmic improvements and expanding the library's feature set. The user's work contributed to the performance and feature enhancement of the zero-allocation hashing library.
hashingopenhftfarmhashzeromurmurhash3
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.