Alexey Lebedev

Working on food delivery platform - Yandex Eats, the biggest food delivery service in Russia, with around 600 thousand orders per day. I'm working on the backend side, developing micro services in logistics domain, mostly focusing on delivery time prediction (ETA) and supply and demand balance (SURGE).

Key results:

• Design and implementation of time prediction algorithms (ETA) using ml and maps services. Improved customer ETA quality, reduced prediction error by 30 percent, reduced number of support requests related to ETA by 50 percents
• Improved courier experience with ETA. Designed and implemented infrastructure for courier timers calculation and fixed bugs. Reduced the number of courier timer delays by 10 percent.
• Optimized requests to maps services. Implemented route caching, reduced number of requests to maps by 50 percent with caching, reduced api latency by 60ms.
• Improved code quality. Made a refactoring, applied design patterns. Helped to onboard new people to work with the codebase.
• Helped to build infrastructure service for calculation delivery offers.
• Implemented availability features of the services for supply and demand calculation
• Led several projects and deliver it in time

Technical Skills:

Proficient in C++ (17, 20), Python, Redis, Postgresql, Messaging Queues, System Design, Distributed Systems

Team Collaboration:

• Collaborated with cross-functional teams, providing support and guidance to new team members.
• Developed conversion and system implementation plans.
• Discussed issues with team members to provide resolution and apply best practices.
• Contributed ideas and suggestions in team meetings and delivered updates on deadlines, designs, and enhancements.

Worked on neural network inference framework OpenVINO (https://github.com/openvinotoolkit/openvino). Contributed to Python API and Benchmarking tool:

• Fixing bugs
• Implemented openvino python api 2.0 with pybind11, improved user experience with api
• Rewrote python benchmarking tool with api 2.0
• Added support of dynamic shape inference in benchmarkig tool
• Collaborated with customers and researched their problem with python GIL, implemented solution
• Contributed to documentation

Technical Skills:

Python, C++, Deep learning, Pybind11, Cython

Worked on neural network inference framework - OpenVINO (https://github.com/openvinotoolkit/openvino). Contributed to python api, fixed bugs, implemented new features in benchmarking tool.