PubMatic
Pune, Maharashtra, India Posted 13h ago
PubMatic is seeking a Principal Engineer for the Machine Learning Platform team to architect and build scalable ML infrastructure, platforms, and engineering capabilities. This hands-on technical leadership role focuses on architecture, complex engineering challenges, reusable platform capabilities, and cross-team influence without direct people-management responsibilities.
Key Responsibilities
Architect and build scalable ML platforms, establishing technical standards and driving end-to-end design and implementation.
Design and optimize Kubernetes-based training infrastructure, shared compute clusters, workload scheduling, and resource utilization.
Improve model training and data-processing performance using parallelization, distributed computing, and efficient data frameworks.
Build MLOps capabilities covering experiment tracking, model versioning, hyperparameter tuning, model registries, CI/CD, and automated deployment.
Develop reusable feature engineering components and self-service workflows that shorten experimentation and productionization cycles.
Engineer automated model validation, regression testing, QA frameworks, drift detection, and production reliability mechanisms.
Enable cloud-ready training pipelines and hybrid-cloud execution with performance, scalability, and cost efficiency in mind.
Advance observability and intelligent automation, including automated RCA, knowledge graphs, and AI-assisted operational workflows.
Requirements
Bachelor's degree in Engineering (CS/IT) or an equivalent degree from a well-known institute or university.
8+ years of software engineering experience with a strong track record as a senior individual contributor delivering complex platforms or distributed systems.
Advanced Python skills and strong software design, debugging, and performance-engineering fundamentals.
Hands-on experience with Kubernetes, Docker, Linux, infrastructure automation, and production operations.
Experience designing large-scale distributed systems, data-processing platforms, or ML training infrastructure.
Strong understanding of MLOps, ML training pipelines, model deployment, lifecycle automation, and CI/CD.
Experience with Apache Spark, Hadoop, or comparable distributed data-processing technologies.
Preferred Qualifications
Advanced degree in Computer Science or related field.
Experience building internal ML platforms used by multiple engineering or data science teams.
Familiarity with LightGBM, XGBoost, PyTorch, TensorFlow, Weights & Biases, or MLflow.
Experience with GPU computing, CUDA, NVIDIA Triton, or optimized model-serving systems.
Experience with AWS, hybrid-cloud infrastructure, feature stores, model registries, and automated retraining.