AI Product Engineer - Agentic AI Platforms (Financial Services)

About this role

Capgemini is at the forefront of Generative AI innovation, helping Financial Services clients industrialize GenAI and Agentic AI platforms at enterprise scale.

We are seeking an experienced and innovative AI Product Engineer – Agentic Platforms to join our Financial Services Artificial Intelligence & Business Lines (FS‑ABL) practice. This role is ideal for a consulting technologist with deep expertise in modern GenAI tooling, agentic system design, and enterprise SDLC, who can partner directly with clients to envision, design, develop, and deploy Agentic AI platforms in regulated environments.

In this role, you will work at the intersection of client advisory, AI product engineering, and delivery execution, helping banks, insurers, and capital markets firms transition from GenAI pilots to production‑grade, governed, multi‑agent systems. You will apply leading GenAI frameworks and LLM platforms — including Anthropic, OpenAI, LangChain, LangGraph, DSPy, and vector databases—while operating across the full Agentic SDLC.

P&C Insurance knowledge and experience is a significant plus. Additionally, familiarity with core insurance platforms like Guidewire, DuckCreek or Majesco will be extremely helpful to succeed in this role.

We are looking for candidates across all levels of experience and expertise - junior through senior level AI Product Engineers.

### Requirements

Client Advisory & Product Vision

• Partner directly with Financial Services clients to identify, prioritize, and shape Agentic AI use cases across customer operations, underwriting, claims, risk, compliance, finance, and technology.

• Lead client workshops to define agent personas, responsibilities, autonomy boundaries, human‑in‑the‑loop checkpoints, and escalation logic.

• Translate evolving business needs into agentic product backlogs, roadmaps, and MVP definitions.

• Support executive conversations around GenAI platform strategy, operating models, vendor selection, and scale‑out approaches.

...