Drimva

American Express

Staff AI Engineer - Personalization, Brand, Communications Tech

New York, NY · Hybrid · Staff · Full-time · Posted Oct 30, 2026

Experience
12+ years building large-scale distributed systems, plus strong experience with LLM systems, agentic workflows or advanced ML infrastructure.
Visa sponsorship
American Express may provide visa sponsorship for certain positions, depending on factors such as business unit requirements, the nature of the position, cost, and applicable laws.

About the job

Leads the architecture, development, and operation of production agentic AI systems and shared capabilities across American Express products.

What you'll do

  • Set technical direction for agentic AI architecture, autonomy, and system design; build and operate production systems.
  • Own shared capabilities such as agent orchestration, retrieval and grounding, LLM infrastructure, evaluation, observability, and safety tooling.
  • Lead design reviews and cross-team delivery, evaluate emerging AI approaches, and mentor engineers.

What they ask for

  • 12+ years building large-scale distributed systems, plus strong experience with LLM systems, agentic workflows, or advanced ML infrastructure.
  • Proven ownership of complex agentic systems spanning multiple teams or products.
  • Strong backend, API, data-pipeline, and cloud-infrastructure fundamentals; deep experience with agentic AI planning, tool use, memory, and evaluation.
  • Experience with workflow engines, asynchronous processing, queues, and streaming systems.

Nice to have

  • Experience building agentic systems in fintech or other regulated industries.
  • Fine-tuning, model optimization, inference pipelines, or meaningful open-source AI contributions.

Pay

$144250 - $256250 annually (base salary)

Tools

Python, Go, TypeScript, REST, gRPC, AWS, GCP, Kubernetes

AI focus

Develops enterprise agentic AI systems and platforms, including planning, tool use, memory, RAG, LLM infrastructure, evaluation, observability, and safety.

Model-written summary · the original posting is authoritative

First seen Oct 28, 2026 · Last verified Nov 5, 2026 (UTC)