Lead Scientist, Data Science

XPO India Shared Services seeks a Lead Scientist, Data Science for its India Data Science team in its LTL business unit. The posting lists Pune, IN, 411014, and also identifies Hyderabad/Pune as the location. It specifies one position and 7+ years of relevant experience. The role combines technical leadership with hands-on development and production deployment of AI solutions, particularly Generative AI and retrieval-augmented generation (RAG).
Responsibilities
Design, build and productionize advanced ML/AI models, including Generative AI and RAG architectures. Lead the full model lifecycle from data preparation and feature engineering through training, evaluation, deployment and monitoring.
Establish and improve MLOps pipelines for ML continuous integration, deployment and monitoring. Apply practices for model governance, reproducibility and scalability, and work with engineering teams to integrate AI solutions into production systems.
Mentor data scientists and ML engineers; work with product and business stakeholders to turn AI concepts into business value; and evaluate emerging AI technologies and frameworks.
Ensure models meet performance, reliability and compliance standards. Develop monitoring for model drift, bias and performance degradation, while promoting responsible AI and security-first practices.
Required expertise includes Generative AI techniques such as LLMs, diffusion models and transformers; RAG pipelines; MLOps and ML engineering; Python; PyTorch or TensorFlow; distributed training; cloud platforms such as AWS, Azure or GCP; and vector databases and retrieval systems such as Pinecone, Weaviate or FAISS. The posting also cites MLflow, Kubeflow and Airflow as examples of orchestration tools.
Generative AIRetrieval-Augmented GenerationLarge Language ModelsDiffusion modelsTransformersMLOpsMachine Learning EngineeringPythonPyTorchTensorFlowDistributed trainingAWSAzureGoogle Cloud PlatformMLflowKubeflowAirflowPineconeWeaviateFAISSFeature engineeringCI/CD for MLModel monitoringModel governanceVector databasesTechnical leadershipMentoring

Work arrangement
No