AI Engineer

Job Description

Title:

AI Engineer (P3)

Location:

Remote - USA, Canada, Europe or Australia

Reports to:

Sr. Director, AI & Innovation

Vacancy

New - Internal Posting

You build the full-stack components that power Innovation Crew pilots and prototypes. Under the technical direction of the Principal AI Engineer, in close collaboration with the AI Experience Engineer, you write the backend services, API integrations, data connectors, and frontend implementations that turn AI architectures into working systems. This role has a high ceiling and a fast runway: you will be working alongside engineers at the frontier of AI application development, with hands-on exposure to frontier model APIs, agentic frameworks, Power Agents library development, and JD Power's core data infrastructure.

You arrive with strong engineering fundamentals and a genuine appetite for AI systems. You gain deep, production-proven agentic engineering experience.

In your first 30 days, you will have your first feature shipped inside a PoC sprint and the codebase oriented. By day 60, you will have your first Power Agents contribution reviewed and merged. By day 90, you will have full implementation ownership of at least one PoC component.

The Innovation Crew builds alongside its own AI agent workforce. As AI Engineer, you will contribute to building and maintaining those agents: writing the tool integrations, evaluation harnesses, and API connectors that give agents access to JD Power's data and systems. You learn to build for observability from day one - every agent pattern you ship is instrumented, testable, and inspectable.

The bar here is not "it worked in the demo" but "it behaves predictably, and we can prove it." When a pattern proves reliable, you contribute it to the Power Agents library, where 400+ engineers & knowledge workers can build on it.

Innovation Crew pilots only get built because the engineering work that underlies them gets done cleanly and quickly. You are the execution engine that turns architectural direction into working code. You build backend services that connect to Snowflake, API integrations that wire agent outputs to JD Power's core platforms, frontend components the AI Experience Engineer builds on. The pilots you help build will be handed off as operational systems to maintaining teams within Product Engineering, Engineering, OEM Solutions, Infrastructure, and Internal Platforms.

You will contribute directly to the Power Agents library, meaning code you write will be used by engineers & knowledge workers across JD Power.

• Build full-stack features and components for Innovation Crew pilots under the technical direction of the Principal AI Engineer: backend services, REST APIs, data pipeline connectors, and frontend implementations; writing clean, well-documented code oriented toward handoff to maintaining teams within Product Engineering, Engineering, OEM Solutions, Infrastructure, or Internal Platforms.
• Integrate Innovation Crew builds with JD Power's core data infrastructure:
  Snowflake, internal APIs, cloud platforms, and the API Gateway following published gateway standards and working with Internal Platforms to confirm data access, schema conventions, and connection patterns before building.
• Contribute to Power Agents library development: implement new agent patterns under the direction of the Principal AI Engineer, write tests, instrument traces, and document behavior so every module you contribute is auditable, testable, and reusable by the engineering organization.
• Contribute to the team's AI agent workforce: build tool integrations, evaluation harnesses, and observability instrumentation that give development agents access to JD Power's systems; implement the baseline evals that let the team verify agent behavior is reliable before relying on it.
• Consume and configure frontier model APIs, agentic frameworks (Lang Graph, CrewAI, or equivalent), and MCP server integrations as directed; develop hands-on proficiency with the team's AI engineering stack in a production-adjacent environment with real delivery…