Search is the core of Yandex. Every query turns into a complex task: to select the best document from billions, one that will truly be useful and answer the person's question. The quality of this solution determines everything: trust in Search, Alice's success, and the future of generative answers in the results.

We are looking for a leader who will take on the development of ranking—one of the largest ML systems in the country—and will be responsible for the quality of Search for millions of users. You will define what Search will become in the coming years: from strategic development directions to specific solutions that will reach production and change the results.

What tasks await you

Develop Search Quality Search is Yandex's key service with a huge audience and business significance. Our goal is to make it the best way to find an answer on the internet.

You will define the ranking development strategy, formulate requirements for Search quality, and launch improvements that change the entire results set. In doing so, we have to compete not only with global search engines but also with vertical services: marketplaces, classifieds, and aggregators.

Build Search for AI Yandex has many generative products: Alice in chat, Alice on Search, and others. Under the hood, they have different pipelines, created by different teams—but all these "Search killers" rely on Search.

Currently, "Search for people" (blue links) and "Search for AI" are one service, although their tasks are different. Ranking is tailored to the behavior of Runet users: priority is given to Russian-language materials convenient for quick familiarization. The average user searching for information about COVID vaccines will not read clinical reports in English, but an LLM will calmly study such documents and provide a detailed answer based on primary sources—this creates a product "wow" effect.

Our goal is to build a separate ranking mode, "Search for AI," which will end-to-end improve the quality of Yandex's generative products.

Leadership and Management You will lead a team of 35+ ML engineers working on Search ranking: set goals, prioritize directions, decompose large initiatives, and bring them to production.

An important part of the role is team development. You will help engineers and team leads grow, contribute to forming a strong engineering culture, and hire new people who will strengthen the direction.

Cross-functions Search is a complex system where ML does not exist separately from other areas. You will work closely with analytics, backend, and product teams: design solutions together, choose optimal approaches, and control the quality of implementation.

Technological Agenda The search industry is undergoing a new stage: neural retrieval, powerful reranker models, and integration into generative systems are changing the architecture of search platforms.

Our task is to develop the ranking system for this new reality: enhance ML approaches, simplify architecture, and make Search faster for experiments and launching new technologies.

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

We expect you to

• Have managed a large ML team
• Have a strong technical background in search or recommendations
• Be able to form a technological strategy
• Have launched major changes in complex systems
• Have hired, developed engineers, and grown leaders

It will be a plus if you

• Have worked with generative systems
• Have developed ML infrastructure or platforms