Who are we looking for

We seek an experienced Machine Learning Operations Engineer (Senior) to transform cutting-edge research into robust, production-ready services for synthetic data generation and to optimize both deep learning and classical ML algorithms (e.g. tree-based models) at enterprise scale (billions of rows). You will build and tune model pipelines end-to-end, ensuring high performance, scalability, and reliability across diverse workloads and dataset sizes.

Key Responsibilities:

Algorithm Optimization & Scaling

• Optimize bottlenecks of the deep generative models to accelerate training and generation of generative models (e.g. transformer, diffusion, GANs).
• Implement distributed training of the models across multi-GPU clusters.
• Optimize distributed training of traditional ML models (e.g. XGBoost, LightGBM, CatBoost) on billion-row datasets.
• Design best practices for memory management to maximize resource utilization (compute and memory), enabling faster training at lower cost.

Data Handling at Scale

• Collaborate with data engineers to design ETL/ELT workflows handling terabyte to petabyte scale tabular and unstructured data.
• Implement scalable feature engineering pipelines using distributed computing frameworks (e.g. Spark, Dask, or Ray).
• Automate data validation (e.g. schema checks, anomaly detection) with rule-based and ML-driven frameworks.

End to end orchestration

• Build ML pipelines that transition research prototypes into reliable production-grade workflow.
• Package models into Docker containers and deploy using Kubernetes.
• Build automated model and data quality monitoring and validation systems to ensure data integrity throughout the pipeline lifecycle.
• Design robust error handling mechanisms, with automatic retries and data recovery in case of pipeline failures.
• Implement logging, monitoring and alerting systems.