RocketHuntSmart job search
Reach out directly about this role
By city
3-5 years
Experience
Full-time
Employment
Hybrid, Onsite
Work Format
Data Science & ML
Specialization
IT & Tech
Industry
Corporation
Company Type
ML Developer for Alice's Security Team
Alice is a multimodal product of the future, where the power of generative models is combined with various sources of external information: web search, image search, information about organizations from Maps, and much more. In addition, there is image and video generation, AI agents, and this list is constantly expanding.
The main task of our team is to make responses safe for the user. We are engaged in data preparation, training of LLM assessors, classifiers, and reward models. Become part of our team to compete on equal terms with world IT leaders and create the products of the future today!
Analytical tasks around ML models We live in an era where the quality of ML systems is largely determined by data. Therefore, a significant part of our work is collecting learns, validating and accepting models, monitoring production, and promptly searching for problems.
ML development For a smart assistant to work truly effectively, it needs additional ML models: they understand user intent, determine the appropriate scenario, and launch the corresponding response mechanism. Therefore, we develop, train, and maintain a whole set of LLM models, regularly updating them and adding new interaction scenarios.
LLM assessors To prepare learns, acceptances, and measure model quality, we have to process huge amounts of data. To do this quickly and efficiently, we create LLM assessors — models that label data with human (and sometimes higher) quality, but work much faster and cost less.
More about ML at Yandex — in the channel Yandex for ML