Hi! I'm looking for an AI researcher for the AI incubator team of a large retailer.
We will be building cool PoCs, agents, and implementing LLMs into real business processes.
💰 Salary: 260-300 K RUB/month
⏰ Schedule: MSK ± 2, full remote
😎 Citizenship/Location: Russia only
🤖 Project until the end of the year + possible extension.
📃Employment: via Individual Entrepreneur contract
Who are we looking for?
We need a hands-on practitioner who will cover the agent architecture and bring models to production:
- Development of agent architecture: managing dialogue state and context, dynamic tool selection (tool calling), configuring autonomy levels and human-in-the-loop scenarios;
- Deep understanding of the ML model lifecycle: from feature preparation to result interpretation;
- Knowledge of methods for combating overfitting, data leakage, regularization; confident algorithm selection for the task;
- Proficiency in mathematical statistics for hypothesis testing, calculating confidence intervals, and independently designing A/B tests;
- Building a metric system: developing offline and online indicators, cross-validation, direct assessment of business impact;
- Continuous monitoring of model drift and data quality in production;
- Experience in at least one of the following areas: classical machine learning; time series and forecasting; recommendation systems, search and ranking; NLP, LLM and generative AI;
- Languages and libraries: Python (pandas/polars, NumPy, scikit-learn, CatBoost/XGBoost/LightGBM, PyTorch), advanced SQL;
- Writing maintainable production code: modular architecture, unit tests, logging, documentation, working with GitFlow and code reviews;
- Building evaluation pipelines (LLM Evals): controlling hallucinations, response relevance, and monitoring system behavior;
- Ensuring security: protection against prompt injection, prevention of personal data leaks (PII/privacy), and agent access rights audit;
- Practical experience with language model APIs, the Hugging Face platform, and orchestration frameworks;
- Ability to estimate the cost, latency, and scalability of LLM solutions;
- Experience in e-commerce, retail, or other product-based high-load systems;
- Infrastructure and orchestration: Airflow, MLflow, Kafka, Redis, Kubernetes;
- Monitoring and observability: Prometheus, Grafana, OpenTelemetry;
- Experience with GPU infrastructure operation and inference optimization;
- Model adaptation through fine-tuning (LoRA/PEFT);
- Experience in designing agentic workflows or multi-agent systems;
- Understanding of information security principles, privacy, and working with personal data.
Why is it cool here?
🚀 We solve non-standard AI incubator tasks (from hypothesis to presentation at the Expert Committee).
📊 Real highload and Big Data in Retail.
Project Tasks
- Development of generative AI models for synthesizing visual and project content based on structured data from the product catalog within the framework of hypothesis testing (PoC) on the corporate platform for evaluating AI initiatives.
📩 How to apply?
Apply here
