Principal Engineer - AI/ML

Job description

Business IntroductionAt GSK, we have bold ambitions for patients, aiming to positively impact the health of 2.5 billion people by the end of the decade. Our R&D focuses on discovering and delivering vaccines and medicines, combining our understanding of the immune system with cutting-edge technology to transform people’s lives. GSK fosters a culture ambitious for patients, accountable for impact, and committed to doing the right thing, making sure that we focus our efforts on accelerating significant assets that meet patients’ needs and have the highest probability of success. We’re uniting science, technology, and talent to get ahead of disease together.Find out more:Our approach to R&DPosition Summary This role is an exciting opportunity to apply your deep expertise in AI/ML engineering to help shape the future of the Global Clinical Operations (GCO) function. You will design, build and operationalize production-grade machine learning solutions on Azure Databricks, MLflow, Unity Catalog and Azure Machine Learning – bringing rigorous engineering discipline to the full model lifecycle from experimentation through to monitored, governed deployment in a regulated environment. An ideal candidate is passionate about engineering robust, scalable AI/ML systems and extensive hands-on experience deploying, scaling, optimizing models in production at enterprise scale. You will work in agile, cross-functional teams alongside Data Scientists, Data Engineers, Clinical Operations stakeholders and partners from the wider Biostatistics, Data Science and R&D community. Responsibilities  Architect and deploy production-grade ML models and GenAI/LLM solutions that address real business problems in Clinical Operations, including trial feasibility, site selection, patient recruitment forecasting and study monitoring. • Own the end-to-end ML lifecycle: data ingestion, feature engineering, versioning, serving and performance monitoring in production. • Build, maintain and improve MLOps pipelines on Azure Databricks using Unity Catalog / MLflow Tracking, MLflow Model Registry and automated CI/CD workflows for continuous training and deployment. Manage online/batch endpoints, compute clusters, batch inference pipelines and model monitoring dashboards. • Implement robust model serving patterns including real-time REST endpoints, batch scoring jobs and feature stores. • Collaborate with Data Scientists to translate experimental notebooks into scalable, maintainable, production-ready code following software engineering best practices. • Partner with Data Engineers to ensure feature pipelines, Delta Lake tables and streaming data feeds meet the latency and quality requirements of production models. • Establish and uphold MLOps best practices: experiment tracking, reproducibility, model versioning, A/B testing, canary deployments, drift detection and automated retraining triggers. • Ensure all AI/ML solutions comply with GSK’s data governance, GxP regulations, data privacy standards and responsible AI principles. • Stay current with the latest ML engineering tooling, GenAI advances and cloud platform capabilities; evaluate and introduce new technologies where they add measurable value. • Mentor junior engineers and data scientists on production ML engineering practices and cloud-native Azure/Databricks tooling. Why You?This role is based in India and is offered as a hybrid position, combining on-site teamwork and focused remote work. We seek people who deliver practical, reliable AI that drives measurable value. You will join a team focused on learning, inclusion and clear communication. This role offers strong technical growth and meaningful impact on how we use AI across the organisation.Basic Qualification We are looking for professionals with the required skills to achieve our goals: • Bachelor’s or master’s degree in computer science, Software Engineering, Mathematics, Statistics or a related quantitative field. • 9-10 years of hands-on ML engineering or MLOps experience, with a proven track record of deploying and operating ML models in enterprise production environments. • API Development: Packaging models into scalable microservices using frameworks like FastAPI, Flask, MLFlow / UC model serving endpoint (real-time/batch) etc. • Deep expertise in Azure Databricks: Spark-based data processing, Delta Lake, Databricks Workflows, Databricks / UC Feature Store and cluster lifecycle management. • Proficient use of MLflow for experiment tracking, model registry, model packaging (ML model format) and programmatic deployment via the MLflow REST API. • Strong Python programming skills; solid proficiency in ML frameworks including Scikit-learn, PyTorch and/or TensorFlow; experience packaging models as reusable libraries or containers. • CI/CD experience: Git (branching strategy, PR-based workflows), Azure DevOps or GitHub Actions pipelines for automated testing, model validation and deployment gating. • Containerization with Docker – building, tagging and deploying ML images; familiarity with orchestration concepts relevant to ML workloads. • Working knowledge of Azure cloud data and compute services: ADLS Gen2, Azure Key Vault, Azure Monitor and cost optimization for ML workloads. • Excellent problem-solving and communication skills – ability to articulate technical decisions and trade-offs to non-technical clinical and business stakeholders. • Motivated to operate independently and leading delivery