Lead Data Scientist -Platform AI Acceleration

The Applied Artificial Intelligence and Machine Learning (Applied AI/ML) team within Infrastructure Platforms is transforming how the firm delivers strategic infrastructure platforms-based solutions—both by applying AI/ML within engineering workflows and by building scalable AI hosting platforms and capabilities for enterprise use. As Lead Data Scientist and Generative Lead within J.P.Morgan, you will operate as a hands-on engineering leader responsible for designing, building, and running production-grade ML and Generative AI services, while setting technical direction that scales across multiple workstreams. You will remain close to the code and architecture decisions, establish delivery and engineering standards, and ensure solutions meet enterprise expectations for security, stability, and operational rigor. The ideal candidate brings a strong foundation in software engineering and AI/ML, along with proven experience leading the development and production operation of AI-enabled systems in secure, enterprise environments. In this role, you will collaborate closely with Infrastructure Platforms AI teams to address priority use cases, design and build services, and promote best practices for scalable, resilient, and secure AI adoption. You will also mentor engineers, contribute to firmwide standards and thought leadership, and help ensure the organization stays at the forefront of AI engineering advancements. Job Responsibilities Analyze large datasets to extract actionable insights and drive data-driven decision-making Evaluate and assist hardening of AI powered use cases on enterprise platforms, defining and applying evals and production drift monitoring, supported by automated data profiling and quality checks (leakage detection, imbalance, missingness) Select and apply models end-to-end across ML, deep learning, and LLM-based approaches, including training, tuning, calibration/thresholding, robustness testing, and structured error/failure-mode analysis. Co-Develop and implement LLM-based, machine learning models and algorithms to solve complex operational challenges. Ship reusable enablement assets for platform users (playbooks, templates, reference implementations) and continuously improve them using feedback loops from production telemetry and incident learnings. Collaborate with wider technology groups for AI driven workflows and use cases, to understand business needs and translate them into technical solutions. Define standards and practices to ensure regulatory and data-privacy considerations are baked into system design and implementation. Required qualifications, capabilities, and skills Post Graduate qualification (Masters or PhD) Data Science, Computer Science, Mathematics. Building and shipping data-driven/AI-enabled production systems, with significant hands-on model development across statistical, classical ML, deep learning, and LLM-based approaches—covering feature/label strategy, training, evaluation, tuning, deployment, and monitoring. Strong grounding in statistics, probability, and experimental design, with the ability to design evaluations, interpret results, and make decisions under uncertainty. Deep hands-on experience with modern ML/DL stacks (e.g., PyTorch and/or TensorFlow, scikit-learn, Hugging Face Transformers). Proven experience with distributed training and scalable model serving, using modern architectures, tools, and frameworks. Hands-on experience deploying and operating models in cloud production environments, including training/tuning workflows, inference operations, monitoring, and performance/cost optimization. Strong technical depth in LLMs/SLMs, including model selection trade-offs (latency/cost/quality), fine-tuning/adaptation where appropriate, and production serving considerations. Hands-on experience designing and operating RAG systems including quality measurement and grounding controls. Strong technical depth in agentic AI systems, including tool/function calling, orchestration patterns, guardrails, structured outputs, and evaluation for reliability and safety. Preferred qualifications, capabilities, and skills Published technical papers, patents, or significant internal publications; conference presentations (speaker/panel) on ML/GenAI/Agentic AI topics. Open-source contributions, including maintaining or meaningfully contributing to ML/GenAI GitHub repositories (libraries, tooling, eval harnesses, MLOps components). Experience with ML accelerators and performance optimization (e.g., GPUs/TPUs), including profiling, distributed training, and inference optimization.