AI Experience Engineer

J.D. Power•Ontario - Remote, ON

1d•$130,000 - $160,000•Remote

About The Position

The AI Experience Engineer is a full-stack role responsible for the complete human-facing layer of AI-powered systems, from data model through backend service to a polished, production-quality interface. This role is part of the Innovation Crew, focusing on designing and building interactions that make agentic AI capabilities understandable, trustworthy, and useful to engineers, analysts, and business stakeholders. The goal is to create inspectable, traceable systems where humans can assess processes, surface relevant information, and course-correct without being in the critical path of every step. Key milestones include auditing the design system and shipping the first agent interaction component within 30 days, developing a working observability interface by day 60, and documenting design system foundations by day 90.

Requirements

Full-stack engineering with frontend depth: ability to build backend services, wire up data sources, and produce a polished production-quality interface within the same sprint.
Production experience with AI-powered systems, including shipping features like streaming output, agent interaction surfaces, human-review workflows, or trace inspection interfaces.
Understanding that LLM systems are non-deterministic and must be designed for regular inspection from the start.
Proficiency in React (or equivalent) and modern design systems, with the ability to both design interactions and implement them.
Ability to move fluidly between designing in Figma and building in code without significant handoff friction.

Nice To Haves

Experience with agent observability tooling (Langfuse, LangSmith, Phoenix/Arize, or equivalent) and designing interfaces that make trace data useful to non-engineers.
Data visualization for non-technical audiences.
Knowledge of WCAG accessibility standards.
Familiarity with the JD Power product ecosystem or user base.
Eval harness design or familiarity with LLM-as-judge patterns.
Understanding of what "quality" means for an agentic system to inform better inspection interfaces.
Strong product intuition: asking "what does the user actually need from this agent?" before deciding how to surface it.

Responsibilities

Own the full-stack design and engineering quality for all Innovation Crew pilots and prototypes, taking concepts from brief or wireframe through backend service, API integration, and polished production-quality interface.
Ensure every deliverable shown to engineering leadership or business stakeholders is visually consistent, accessible, and immediately legible to non-technical audiences.
Design the human-on-the-loop interaction patterns for agentic systems, building interfaces for engineers and reviewers to inspect agent traces, audit reasoning outputs, adjust parameters, and surface anomalies.
Build and maintain a shared design system and component library for Innovation Crew outputs, including AI-specific components like streaming output displays, agent status and progress indicators, trace inspection views, confidence surfaces, and human-review interfaces.
Define the UX standards and interaction vocabulary for the AI Assistants and Copilots capability pillar, including contextual AI assistance interfaces, real-time knowledge retrieval surfaces, AI-generated content displays, and role-specific intelligence dashboards.
Conduct lightweight user research and usability testing with internal stakeholders and business unit partners, integrating user feedback into the design process.
Partner with the Principal AI Engineer to design evaluation interfaces and agent observability dashboards.
Champion WCAG accessibility standards and performance engineering across all Innovation Crew deliverables.
Apply JD Power brand guidelines to any output shown externally.
Contribute UX and frontend technology evaluations to the Emerging Technology Radar process.
Propose directional Architecture Decision Records for AI interaction patterns adopted by the Innovation Crew.