C++ ML Developer for Yandex Visual Search Team

ML Developer in C++ for Yandex Visual Search Team

Visual Search Team develops CV technologies in the very heart of Yandex. Our algorithms and services are integrated into the company's key products and help millions of users solve everyday tasks. You have definitely met us if you've used image search, Alice, the smart camera, Market, or other services where the image plays the main role in their scenarios.

We not only create innovative scenarios from scratch but also constantly improve existing ones, ensuring their stability and the highest quality at the world standard level.

We are looking for an experienced C++ developer with deep knowledge in ML, who is ready to create and implement complex computer vision technologies in production.

What tasks await you

Designing ML system architecture You will not just write code, but design comprehensive systems for visual search and image analysis. You will be responsible for selecting technical approaches, building pipelines, and ensuring the scalability and fault tolerance of our solutions.

Implementing and optimizing models in production Your key task is to turn state-of-the-art models (CV, LLM/VLM) into fast and stable components of Yandex's high-load services. This includes deep work with C++ code, its profiling, and performance optimization to achieve maximum efficiency.

Participation in the full product life cycle You will not just be an executor, but a full-fledged participant in the process: from generating and testing hypotheses together with the team to rolling out features, participating in PR activities, and analyzing results after launch.

Interaction with R&D and ML researchers You will closely work with Yandex's research teams to quickly adapt the most advanced technologies and implement them into products.

More about ML at Yandex — on the channel Yandex for ML

We expect that you

• Are excellent at C++ and have worked on commercial development with it
• Deeply understand the principles of machine learning and classical algorithms
• Have implemented ML models in production: know how to make a model work fast and stably under load
• Can understand complex codebases, refactor, and improve existing solutions
• Are ready to quickly dive into the structure of Yandex services and find the best points of application for ML technologies

Will be a plus

• Have worked with Python for ML tasks (PyTorch, TensorFlow)
• Understand the architecture of modern neural networks: transformers, VLM, CNN