Vyacheslav Petrov — CV

Backend Tech Lead. 11 years of experience building scalable, high-throughput backend services.

Skills

Negotiated directly with counterparties’ legal counsel and rights holders, carefully managing legal risks in communications and avoiding unintended admissions or commitments.
Designed and implemented backend service architectures, focusing on scalability, performance, and maintainability.
Scale at VK Music: 400+ pods and 400,000+ requests per second. Log ingestion throughput on VK’s infrastructure team: 12 Gbit/s.
Led backend task planning and decomposition, translating technical requirements into actionable tasks for the development team.
Drove projects from planning to delivery, tracking progress, managing technical risks, and ensuring timely completion.
Optimized resource utilization and infrastructure costs, improving CPU, memory, and database efficiency.
Made ClickHouse queries 20× faster on band.link.
Reduced Radio’s pod count by 75% through request pipeline optimizations.
Reduced network traffic on content servers from 700 to 350 Gbit/s through improved CDN caching.
Improved service security, reliability, and fault tolerance through architectural improvements and infrastructure hardening.
Handled critical production incidents and contributed to emergency service recovery.
Led 200+ incident reviews at VK and helped restore Yandex Music and Yandex Radio following a military attack on three Yandex data centers.

2025 – present Backend Tech Lead at radio.yandex.ru.
I designed the backend, and a small team of developers implemented it. The project took about two months from start to finish.

Radio is you.
A project manager
2022 – 2025: Yandex Worked on the stability and observability of band.link.
2019 – 2022: VK Worked on the infrastructure team, improving site performance and content delivery times, and building tools for debugging and system monitoring.
2017 – 2019: VK Worked on the VK Music team. In collaboration with UMA, I built backend systems for importing audio track licenses, automatically resolving licensing conflicts, and performing real-time monetization calculations. I also worked on matching user-generated audio to licensed tracks. The most challenging part was matching tracks modified by pirates in ways that were imperceptible to listeners but affected automated matching.
2016 – 2017: VK Built the backend for a stories service similar to Instagram Stories and Snapchat Stories.

I implemented a system for matching user-uploaded books against rights holders' content. My work was covered in the media and cited as a benchmark for matching quality. The number of DMCA claims dropped from more than 24,000 to 648 per month.
Worked on video content matching.

I rarely check my LinkedIn profile because of the volume of spam. Please don't spam me.