Senior AI Engineer I

As a Senior AI Engineer I – Agentic AI, you will be a core builder responsible for turning complex, ambiguous problems into production-grade agentic systems that operate on real financial data, serve real customers, and meet real regulatory requirements.

You will work end to end: shaping solutions with product and design, building and shipping production code, and owning what you deliver after launch. The scope of this role spans customer-facing LLM-powered features, agentic systems that automate financial workflows, and internal AI capabilities that enable other engineers to build with AI safely and efficiently.

This is not a research-only role. We are looking for engineers who are comfortable operating with autonomy, exercising sound judgment, and pushing the technical envelope within the realities of a regulated financial environment.

What You’ll Do

• Design, build, and ship LLM-powered and agentic product features that change how customers manage their finances.

• Build agentic AI systems that reason over context, invoke tools, take real actions, and recover gracefully from failure.

• Architect and implement production-grade RAG pipelines over sensitive financial data, with strict requirements for correctness, auditability, and safety.

• Contribute to shared AI infrastructure, including LLM services, agent orchestration frameworks, and evaluation and monitoring tooling, that scales agentic development across Amex Technology.

• Own the systems you build in production, including reliability, latency, cost, and failure modes.

• Work closely with product and design partners; engineers in this role are expected to think in terms of customer outcomes, not just technical execution.

• Technical Environment: We don’t hire to a narrow checklist, but candidates should be comfortable operating in a modern, enterprise-scale environment with a strong emphasis on agentic AI.

• Core engineering stack: Languages: Python, Go, TypeScript.

• Cloud and infrastructure: AWS and/or GCP, Kubernetes.

• APIs and services: REST, gRPC.

• Distributed systems: event-driven architectures, including Kafka.

• Agentic AI and ML: Commercial and open-source LLMs integrated into agentic workflows.

• Tooling for agent orchestration, retrieval-augmented generation, vector storage, and evaluation.

• Strong schema, validation, and state management practices.

• AI-assisted development: Fluency with AI-assisted and agentic development workflows for design, implementation, testing, debugging, and refactoring.

• Thoughtful use of these tools while maintaining production-quality engineering standards All systems are built to meet high standards for reliability, security, and auditability, reflecting the responsibility of deploying autonomous AI in a financial services environment.