Machine Learning Engineer (Federated & Edge AI) (вакансія неактивна)

About the Role

We’re looking for an experienced Machine Learning Engineer with strong expertise in federated learning, distributed training, and AI system orchestration.
You’ll be responsible for designing, implementing, and deploying ML pipelines for cross-silo federated learning environments using FedML, Docker, and GPU-based infrastructure.
Your work will directly shape how distributed AI systems train collaboratively without exchanging raw data, ensuring performance, privacy, and scalability.

Responsibilities

• Develop and optimize federated learning pipelines that coordinate multi-client model training using FedML (e.g., FedAvg, cross-silo setups).
• Design and maintain Docker-based infrastructure for running master and client containers with GPU access.
• Develop and expose APIs for model orchestration, checkpointing, and evaluation.
• Design and run A/B experiments to compare training setups and federated strategies.
• Collaborate closely with DevOps, Data, and Product teams to ensure scalable, reproducible, and secure deployments.

What We Expect from You

• 2+ years of experience as a Machine Learning Engineer or MLOps Engineer.
• Strong skills in Python, PyTorch or TensorFlow, and SQL.
• Proven experience with Docker, Docker Compose, and container orchestration.
• Experience with FedML, Federated Learning frameworks, or distributed ML systems.
• Ability to create end-to-end ML pipelines, from data to deployed models.
• Good documentation practices and comfort with research-oriented spike projects.

Nice to Have

• Knowledge of federated data privacy techniques (e.g., differential privacy, secure aggregation).
• Experience with Edge AI, on-device learning, or IoT-oriented deployments.
• Contribution to open-source ML frameworks or academic ML research.