This is a hybrid role: you will be both an AI Lead (hypothesis development, working with contractors) and a Data Scientist (models, data, ML).
Responsibilities
🔬 1. Business Hypotheses
- Interact with business stakeholders (marketing, sales, product, HR, logistics), identify AI application points
- Formulate and test hypotheses: "What if we automate / predict / cluster / recommend this?"
- Evaluate hypothesis effectiveness (development cost vs. benefit)
🤖 2. Working with LLMs and YandexGPT
- Develop the current integration with LLM (Yandex) — we are already using it, need to scale
- Design RAG (Retrieval-Augmented Generation) solutions for corporate data responses
- Develop AI assistants for: Ad-hoc data analysis requests — answering typical data questions instead of creating typical dashboards, internal processes — document search, report generation, indicator interpretation
📊 3. Data Science (models and data)
- Develop ML models (classification, regression, clustering, time series forecasting)
- Collect and prepare data for models from DWH and other corporate sources
- Conduct EDA, feature engineering, validation, A/B tests
- Confirm model quality — formal quality metrics (business metrics, not just ML metrics)
- Document results (reports, presentations for stakeholders)
🤝 4. Working with contractors and pilot projects
- Search the market for external AI solutions that meet business needs
- Formulate technical specifications for contractors, evaluate their competencies
- Manage pilot projects: deadlines, KPIs, acceptance of results
- Prepare factual basis for decision-making: build (do it ourselves) vs. buy (buy a ready-made solution) vs. partner (hire a contractor)
- Current pool: LLM projects with Yandex; plans include — AI assistants for data, NLP solutions, Computer Vision (if needed)
🛠️ 5. Integration and Implementation
- Bring models to production — integrate them into the existing DWH infrastructure
- Write ETL/ELT pipelines for ML (jointly with DWH engineers)
- Monitor model quality in the production environment, set up auto-retraining
- Integrate AI solutions with the company's current systems (CRM, BI, corporate portal)
Requirements
- Experience with complex queries, window functions, CTEs, DWH (PostgreSQL / ClickHouse);
- Experience with YandexGPT / GPT / Open-source LLM, understanding of RAG, prompt engineering;
- Possess basic development skills;
- Have experience managing projects from requirements gathering, defending ideas to stakeholders, to presenting results.
Also:
- Pandas, NumPy, Scikit-learn, FastAPI / Flask;
- Gradient boosting (CatBoost, XGBoost, LightGBM), linear models, clustering, metrics, validation.
Conditions
What we offer:
- Work in one of the largest international companies, with brands that are truly loved;
- Competitive salary (based on interview results), annual bonus;
- Official employment, compliance with the Labor Code of the Russian Federation;
- Paid sick leave up to 100% of salary;
- Voluntary medical insurance policy (extended with dentistry), life insurance;
- Meal compensation;
- "Ponimayu" service — free consultations with psychologists, lawyers, financial advisors, healthy lifestyle specialists;
- Five-day work week (hybrid work format), Friday is a shortened working day (until 15:30);
- Office in the center of Moscow (7 minutes from Nov Slobodskaya and Mayakovskaya metro stations);
- Coffee sets as a gift every quarter;
- The best coffee in the office :)
