The Unified Advertising Database is a relational database built on top of YTsaurus dynamic tables. Our database stores, processes, and supplies advertising data to other Yandex services. It combines all the capabilities of YTsaurus dynamic tables (transactionality, scalability) and stream processing. This allows us to flexibly combine transactional and non-transactional approaches when solving problems.

The goal of our project is to build a database that can replace traditional relational DBs. The project is at the heart of rapidly evolving advertising technologies, and the infrastructure solutions we create are subsequently used across all of Yandex.

What tasks await you

Database Core You will be optimizing query execution speed and the storage itself using more compact data structures. At the same time, you will have the opportunity to contribute to the YTSaurus codebase.

Processing Platform You will be tasked with developing our own framework for stream data processing: adding new features to replace the MapReduce data processing paradigm with streaming.

Core Infrastructure We are constantly improving the stability and increasing the speed of our services. You will also participate in this.

We expect that you

• Are ready to develop in C++
• Know classic algorithms and data structures
• Understand the concurrency paradigm, are familiar with main algorithms and common problems in this area
• Understand key aspects of building fault-tolerant distributed systems
• Can write easily readable code

Will be a plus

• Understand database principles
• Have developed distributed data storage and processing systems