Vlad Sterzhanov is a Staff Software Knight with 13 years of experience building large-scale systems that push healthcare from reactive to preventative, currently developing infrastructure at GRAIL to detect cancer early and extend healthy human lifespan. He blends deep ML and distributed-systems expertise—contributions include neural-network tooling for Yandex’s ML repo and protocol-level improvements to Freenet’s daemon—demonstrating both algorithmic and low-level networking chops. His career spans major product and research environments (GRAIL, Asana, Facebook, Mozilla) where he ships robust, production-ready infrastructure and mentors engineering practice. Trained at MIPT and the Yandex School of Data Analysis, he pairs rigorous mathematical grounding with practical software engineering. Based in San Francisco, he favors projects that translate complex research into reliable, impactful systems. Colleagues describe him as a pragmatic engineer who routinely tackles hard, non-obvious bottlenecks in both ML pipelines and networked systems.
13 years of coding experience
6 years of employment as a software developer
Belarusian State University
Bachelor of Science (BS), Mathematics and Computer Science, Bachelor of Science (BS), Mathematics and Computer Science at Moscow Institute of Physics and Technology (State University) (MIPT)
Free Master’s-level program, Department of Big Data, Free Master’s-level program, Department of Big Data at Yandex School of Data Analysis
High School, Physics and Mathematics, High School, Physics and Mathematics at Lyceum of Belarusian State University
Contributions summary:Vlad focused on implementing and refining a machine learning toolbox. They initialized base classes for regression and classification, including parameters for different network types. The commits demonstrate the development of neural network models and their integration within the toolbox, along with fixes related to deepcopy and model scaling.
Contributions summary:Vlad primarily focused on enhancing the core functionality of the Freenet REference Daemon, specifically addressing bandwidth limitations and packet handling. They implemented cumulative acknowledgements and reworked the cummack structure, introducing distant ranges for improved efficiency. The user's contributions also included backward compatibility measures and the modification of several core Java files related to packet sending and data transmission.
distributeddaemonp2pfreenet
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.