ml engineer in banking

Описания нет

• Build and evaluate classical machine learning and deep learning models against baselines
• Build LLM-based solutions using prompting, RAG, structured outputs, tool calling, and agentic workflows; apply LoRA/QLoRA fine-tuning when required
• Design end-to-end ML system architectures for batch and real-time inference and scalability
• Develop models as APIs or services and integrate them with banking systems
• Deploy solutions in on-premise or private cloud environments
• Establish CI/CD, model registries, drift and quality monitoring; update and support production models
• Assess model explainability and fairness, document solutions with model cards and audit trails, and ensure regulatory compliance with Risk/Compliance

• At least 2 years of hands-on experience in ML/AI or Data Science
• Experience delivering ML or GenAI solutions to production end-to-end, not only PoCs
• Higher education in Artificial Intelligence, Computer Science, Mathematics, Statistics, or a related field, or equivalent experience
• Experience with supervised and unsupervised classical ML, feature engineering, cross-validation, hyperparameter optimization, business-driven metric and threshold selection, A/B testing, and statistical fundamentals
• Experience developing and training deep learning models with PyTorch, including MLP, CNN, RNN/LSTM, and Transformer architectures; ability to choose between Classical ML and DL and optimize inference
• Knowledge of prompting, embeddings, RAG, Vector DB, hybrid search, reranking, structured outputs, tool calling, agent frameworks, PEFT fine-tuning with LoRA/QLoRA, and LLM evaluation and safety
• Knowledge of ML system design and MLOps, including batch/real-time inference, microservices, event-driven architecture, API design, model versioning, latency-accuracy-cost trade-offs, MLflow or similar, experiment tracking, drift monitoring, and CI/CD
• Strong Python skills and SQL for large-scale data; experience with Git, REST APIs (FastAPI), Docker, and unit/integration testing
• Understanding of working with confidential and regulated data in isolated/on-premise environments
• Ability to clearly explain technical topics to non-technical audiences

• Location: Baku, Baku Economic Zone, Azerbaijan