Production-Grade AI & LLM Director

Testing Site - JPMC; Greater London; On-site; GBP 150,000 - 230,000; Full time

Job summary

Testing Site - JPMC is seeking a Generative AI Executive Director in the Chief Analytics Office to lead the design and delivery of production-grade LLM systems powering mission-critical products. You will bridge AI research with engineering, work with ML, product, and cloud teams, and drive scalable, secure AI solutions at scale.

In this role, you will architect APIs and agentic workflows, own end-to-end delivery, and influence strategy for enterprise AI across thousands of professionals.

Qualifications

• PhD or equivalent in CS/Math/Stats or related field.
• Hands-on shipping production AI systems.
• Expertise in NLP, CV, or multimodal LLMs; statistics and ML theory.
• Experience with distributed, multi-threaded, scalable apps (Ray/Horovod/DeepSpeed).
• Strong ability to communicate complex concepts to stakeholders.

Responsibilities

• Architect scalable APIs and agentic workflows across the firm.
• Deliver production LLM-based systems for text, image, speech, and video applications.
• Own end-to-end delivery, performance, and improvement of LLM Suite products.
• Collaborate with ML engineering, product mgmt, and cloud engineering to ensure reliability and security.
• Bridge AI research with engineering for production-ready solutions.
• Drive results with an entrepreneurial mindset in a fast-paced, high-impact environment.

Skills

ML engineering

NLP & multimodal LLMs

Architecting scalable APIs

Stakeholder communication

Education

PhD or equivalent in CS/Math/Stats

Tools

Ray

Horovod

DeepSpeed

Job description

Testing Site - JPMC is seeking a Generative AI Executive Director in the Chief Analytics Office to lead the design and delivery of production-grade LLM systems powering mission-critical products. You will bridge AI research with engineering, work with ML, product, and cloud teams, and drive scalable, secure AI solutions at scale.

In this role, you will architect APIs and agentic workflows, own end-to-end delivery, and influence strategy for enterprise AI across thousands of professionals.