Senior ML Engineer (Banking) in Oschadbank

MY BANK IS MY SUPPORT. 400">MY PEOPLE.  MY BUSINESS. MY FREEDOM.

Key tasks:

• development, training and validation of ML/DL models for key banking use cases;
• building and optimization of feature engineering pipelines for big tabular data and unstructured documents;
• setting up hyperparametric search (Optuna / Ray Tune), reproducible experiments;
• preparation of models for production: inference optimization, formatting (ONNX/torchscript), integration with MLOps;
• participation in the design and implementation of monitoring pipelines, drift detection, retraining;
• cooperation with Data Engineers (data pipelines, feature store) and DevOps (CI/CD, infrastructure);
• mentoring ML and DS teams, participation in code reviews, writing technical documentation.

Technical requirements:

1.       Machine Learning & Deep Learning:

• 5+ years of experience in ML/DL, of which 2+ years in production;
• PyTorch: building models, custom layers/losses, basic distributed training;
• hands-on experience with XGBoost / CatBoost / LightGBM in production;
• feature engineering and feature selection for high-dimensional tables data;
• hyperparameter optimization (Optuna, Ray Tune);
• basic experience in model monitoring, drift detection, retraining.

2.       MLOps & Infrastructure (at the level of confident cooperation):

• understanding of Docker/Kubernetes for containerization and orchestration of ML services;
• experience with CI/CD for ML (GitLab CI / GitHub Actions / Jenkins);
• knowledge of model versioning and experiment tracking tools (MLflow, DVC, W&B);
• preferred: experience with feature store (Feast / analogues), model serving (TorchServe, Triton, ONNX Runtime);
• cloud or on-prem ML stack (AWS SageMaker, Azure ML, GCP Vertex AI or internal solutions).

3.       Software Engineering:

• Python 3.10+ (strong level): typing, modular architecture, clean code;
• REST API (FastAPI), Pydantic, basic asynchronous stack (asyncio / aiohttp);
• work with SQL (PostgreSQL, Oracle) and NoSQL/columnar (MongoDB, ClickHouse);
• Git, code review, unit/integration tests (pytest / unittest).

4.       Production & Performance:

• understanding of model optimization: quantization, pruning, distillation (as a plus);
• Trade-offs batch vs real-time inference, latency / throughput / cost;
• basic experience with distributed training (PyTorch DDP / Horovod);
• monitoring and logging (Prometheus, Grafana, ELK or analogues).

Bank specifics

1. Regulatory & Compliance:

• practical experience explainability (SHAP, LIME, feature importance);
• participation in model validation / model risk management processes (backtesting, stability).

2. Banking Use Cases (experience of at least 2-3):

• Credit scoring / credit risk;
• Fraud detection (online/near real-time);
• AML / KYC;
• Churn prediction, next-best-offer / next-best-action;
• Collections / collection strategies;
• Credit Memo.

3. Data Security:

• work with PII, anonymization/masking;
• understanding of secure deployment, encryption, access (OAuth2, JWT, RBAC – as a plus).

Soft Skills & Leadership

• experience working in cross-functional teams (DS, DevOps, Risk, Business);
• mentoring junior/middle ML/DS specialists;
• ability to explain model results to non-technical stakeholders;
• active participation in code review, standardization of modeling approaches;
• experience working for Agile/Scrum.

We guarantee:

• conditions for professional and career growth of each employee;
• official employment with compliance with all social guarantees according to the Labor Code (paid sick leave, vacation from 29 calendar days per year);
• free health insurance for all employees.

Send your resume - become a partteamSAVE!

Save in on social networks:

Facebook       https://www.facebook.com/oschadbank/ 

Linkedin          https://www.linkedin.com/company/oschadbank/  

Instagram      https://www.instagram.com/oschadbank_official/ 

Tiktok                https://www.tiktok.com/@oschad