Vacancy: Data / ML Platform Engineer (Middle+ / Senior) Company: AstraTech Format: hybrid / remote within Russia Location: Moscow Employment: full-time Salary Range: 280,000 - 400,000 ₽ gross

A HealthTech startup is creating a medical decision support product that makes healthcare processes more efficient and helps doctors make more accurate decisions 🕊

The team is currently building a Data Platform and ML / LLM Platform. This is a unified data pipeline: from loading from various sources to the measured model result. The team is looking for an engineer who will take responsibility for data preparation and tools for the ML team, and who can conduct model experiments independently.

Your responsibilities will include: — Developing Python / SQL pipelines for loading and processing data from files, APIs, and information systems. — Designing storage tables: entities, relationships, transformation rules. — Integrating pseudo-anonymization and quality checks into data streams. — Preparing features and versioned train/validation/test datasets. — Ensuring traceability and reproducibility of ML / LLM experiments. — Collaborating with the ML team to develop model evaluation, data preparation for RAG, and fine-tuning.

We expect from you: — Proficient Python and SQL, maintainable code beyond notebooks. — Practical experience with ETL / ELT: orchestration, incremental processing, retries. — Experience with S3, Parquet, pandas / Polars / PyArrow. — Understanding of the ML lifecycle: datasets, baselines, metrics, leakage checks. — Experience with MLflow, ClearML, or similar tools; Linux, Git, Docker, CI/CD. — Ability to work with sensitive data.

Will be a plus: — Experience with medical data, EMR systems, and clinical reference data. — Practical NLP / LLM: data extraction from text, embeddings, RAG. — PyTorch / Hugging Face, fine-tuning, and inference on GPU. — Experience transforming a research prototype into a regular pipeline.

Why it's interesting: — A project that genuinely benefits people. — Opportunity to build a platform from scratch and influence its architecture. — Work at the intersection of Data Engineering, MLOps, and LLM with a modern tech stack.