Wisebits — is an international IT holding with a strong technical team! Our key product is a high-load video hosting platform used by millions of people worldwide every day. With over 15 years on the market, we’ve built a professional team that values engagement and initiative.

If you’re excited about working with a high-load product and want to have an impact on its development, join us!

What you'll be working on:

• Participating in the development and improvement of recommendation systems, including user-item, item-item, and cross-platform recommendations.
• Optimizing recommendations on category, channel, and tag pages.
• Building text-to-video search: embedding models, ranking, and semantic search.
• Turning ML prototypes into production services: inference optimization, profiling, feature preparation, and model quality control.
• Developing high-load ML services in Python, including gRPC/Triton-based services.

Our stack:

• Languages and libraries: Python, PyTorch, CatBoost, NumPy, Pandas, Polars, CuDF, CuPy.
• Infrastructure and deployment: Kubernetes, Docker, Airflow, Triton Inference Server.
• Data storage: ClickHouse, MySQL.

What we expect from you:

• 2+ years of experience with recommendation systems, search, or related ML tasks.
• Strong knowledge of Python and ML frameworks, such as PyTorch, CatBoost, etc.
• Experience deploying models to production.
• Understanding of offline and online model evaluation principles.

What we offer: