Team Lead for the Data Science team of the ML platform for defining ad parameters

About the team

Avito hosts hundreds of millions of listings. To make them as useful as possible for users, we determine the parameters of goods and services based on text and photos. For this, our item2param team develops a platform where any Avito Data Scientist can quickly train and put an ML model into production — for example, for automatically determining a listing's category, job title in vacancies, or renovation quality from a photo. More than 1000 models are already running on the platform and helping in numerous Avito business scenarios.

Our goal is to scale the platform so that it can automatically determine all parameters of all listings, as well as lower the entry barrier so that even employees without ML experience can create and launch models.

We are looking for a strong DS Team Lead to head the team and take the platform to the next level.

Examples of future tasks:

•  Manage a team of five DS engineers: planning hiring, interviewing candidates;
  

•  Develop and mentor the team: assisting colleagues' professional growth, developing expertise within the team, conducting performance reviews;
  

•  Plan, set goals (quarterly and annual) and achieve them through the team's coordinated work;
  

•  Take responsibility for the team's technical decisions: designing service architecture, increasing the level of their technical maturity;
  

•  Automate training: setting up pipelines for data collection and labeling (including with the help of LLMs), retraining and monitoring model quality;
  

•  Scale and optimize the service for growing load (thousands of requests per second) and new models;
  

•  Improve the quality of current base models (text and image transformers);
  

•  Improve the quality of individual scenarios — for example, category prediction or named entity recognition (NER).
  

We expect you to:

•  Have at least one year of experience managing a DS team;
  

•  Have good knowledge of classical ML and experience in DL tasks for texts and images;
  

•  Have at least four years of experience implementing and operating ML solutions;
  

•  Be proficient in Python and major ML frameworks (Torch, Pandas);
  

•  Understand MLOps principles and know how to use tools for automating training, deployment, and monitoring of models.
  

Working with us is:

•  An opportunity to improve the experience of millions of users;
  

•  Interesting and complex tasks on a large scale;
  

•  A strong and very friendly team of DS and engineers;
  

•  An opportunity to learn and try new things, with powerful hardware for it;
  

•  A training budget that can be spent on courses or professional literature;
  

•  Health care: from day one you will have voluntary health insurance including dentistry, with a general practitioner and massage therapist available at the office;
  

•  The possibility to work remotely or from offices in four cities in Russia.