Applied Scientist / MLOps Engineer

Location: Russian Federation Format: Remote Period: Long-term Employment: Full-time Contract type: Agreement with sole proprietor (IP) Compensation: from 1200 rubles per hour to 1500 rubles per hour

Applied Scientist / MLOps Engineer (Data Anonymization / Privacy AI)

Responsibilities (Key Tasks): 1. Development of a data classifier for DB fields (Tabular ML) Development of an existing ML model (RandomForest) that identifies personal data in database fields based on metadata and values. Training the model to recognize new data types (e.g., VIN numbers, specific identifiers). Improving classification quality: fixing errors (e.g., detecting complex phone number formats). Solving the problem of "mixed data": training the model to work correctly when a DB column contains a mix of full names, addresses, and phone numbers. Specifics: Optimizing the model for high-speed CPU operation, as the system will need to perform a full scan of databases several TB in size. 2. Creation of a model for identifying data in arbitrary text (NLP / NER) Developing the architecture and training an NER model (Named Entity Recognition) for finding personal data in unstructured text. Working with various file formats: TXT, DOCX, PDF (including text recognition) and logs. Implementing pipelines for cleaning and normalizing "noisy" data before feeding it into the model.

What we expect from you (Requirements): Experience and Leadership: 3+ years of experience in developing and implementing ML models in production. Experience managing the full cycle: from research and task definition to quality monitoring and setting up automatic retraining. Leadership: Readiness to lead the direction, independently decompose business tasks into technical ones, and be responsible for the result. Technical Stack (Python and ML): Proficient in Python (pandas, numpy, scikit-learn). Deep understanding of classical ML algorithms (decision trees, ensembles, Random Forest), ability to interpret results (Feature Importance) and tune hyperparameters. NLP: Practical experience with Hugging Face Transformers and/or spaCy. Experience training NER models is mandatory. Data Engineering: Experience in building unstructured data processing pipelines (PDF, texts). Feature Engineering skills for tabular data. Performance and Infrastructure (MLOps): Performance: Experience optimizing model inference for CPU operation. Scale: Experience working with large data volumes (full database scans from several TB). Production: Experience deploying models via API (FastAPI), containerization (Docker). MLOps: Proficiency in versioning tools (MLflow, DVC), understanding of CI/CD processes. Experience in setting up model quality monitoring (Model Drift, Quality Gates) and organizing automatic retraining.

Will be a plus: Experience specifically with classifiers for DB data profiling. Experience in recognizing specific data types (VIN, card numbers, crypto addresses). Experience in personal data protection, data anonymization, or anti-fraud systems.

What we offer: Opportunity to influence the architecture of a complex AI product from scratch. Work with real Big Data tasks (TB+). Modern tech stack and absence of legacy code.

Contact: recruiter contact