About the Role
Join the Chief Data and Analytics Office (CDAO) organization at J.P. Morgan, the driving force behind the firm-wide adoption of Artificial Intelligence (AI). This team oversees data use, governance, and controls for the build, adoption, and maintenance of cloud infrastructure, data, and AI/ML products, ensuring both effectiveness and responsibility.
As a Generative AI Executive Director within the CDAO organization, you will be instrumental in the smooth operation and optimization of LLM-aided AI products. The firm-wide team focuses on developing scalable LLM-based products and reusable back-end APIs. You will collaborate closely with cross-functional teams, including the ML Centre of Excellence, AI Research, and Cloud Engineering, to foster innovation and deliver high-ROI solutions. A key focus will be designing scalable APIs with clear separation of concerns and well-defined interfaces, enabling other teams to build their own ML products and solutions.
Job Responsibilities
- Combine vast data assets with cutting-edge AI, including LLMs and Multimodal LLMs.
- Bridge scientific research and software engineering, requiring expertise in both domains.
- Collaborate closely with cloud and SRE teams while leading the design and delivery of production architectures.
Required Qualifications, Capabilities, And Skills
- PhD in a quantitative discipline (e.g., Computer Science, Mathematics, Statistics).
- Experience in an individual contributor role in ML engineering.
- Proven track record in building and leading teams of experienced ML engineers/scientists.
- Solid understanding of the fundamentals of statistics, optimization, and ML theory, with a focus on NLP and/or Computer Vision algorithms.
- Hands-on experience in implementing distributed/multi-threaded/scalable applications (including frameworks such as Ray, Horovod, DeepSpeed, etc.).
- Ability to understand and align with business expectations, and write clear and concise OKRs (Objectives and Key Results).
- Experience as a "Responsible Owner" for ML services in enterprise environments.
- Excellent grasp of computer science fundamentals and SDLC best practices.
- Ability to understand business objectives and align ML problem definition.
- Strong communication skills to effectively convey technical information and ideas at all levels, building trust with stakeholders.
Preferred Qualifications, Capabilities, And Skills
- Experience in designing and implementing pipelines using DAGs (e.g., Kubeflow, DVC, Ray).
- Ability to construct batch and streaming microservices exposed as gRPC and/or GraphQL endpoints.
- Demonstrable experience in parameter-efficient fine-tuning, model quantization, and quantization-aware fine-tuning of LLM models.
- Hands-on knowledge of Chain-of-Thoughts, Tree-of-Thoughts, Graph-of-Thoughts prompting strategies.