Staff AI Designer

Intercom is a leading AI Customer Agent company focused on enhancing customer experiences through advanced AI solutions. The Staff AI Product Designer will define AI capabilities, design AI behaviors, and shape human-AI workflows, ensuring reliable and high-quality user experiences powered by AI.

Responsibilities:

• Define AI capabilities (the “what and why”)
• Identify and pattern match high-value opportunities for AI across the product
• Define what AI should do—and just as importantly, what it should not do, making hard or unpopular calls when needed
• Set clear boundaries for autonomy vs human control
• Design AI behavior, orchestration, and system logic
• Architect AI-driven workflows and agentic systems that operate across real-world scenarios and over time, including multi-step and agent-to-agent interactions
• Establish decision frameworks: when systems should act, ask, escalate, or defer
• Design for uncertainty, failures, and edge cases as core system behavior
• Develop and apply design principles and heuristics to shape how AI systems behave and make decisions
• Ensure systems remain coherent and predictable, even as complexity scales
• Shape human + AI workflows and end-to-end experience
• Design experiences for users who are increasingly managing AI systems and agents
• Design how AI integrates into real user workflows, not just isolated interactions
• Define the role of the human in the loop: where users guide, review, or override
• Ensure users can understand, trust, and recover from system failures
• Partner with product designers to continuously shape and adapt the product experience as underlying systems evolve
• Define quality, evaluation, and reliability
• Define what “good” looks like (e.g. trust, accuracy, effort reduction, and whether outputs are structured and usable in context)
• Design evaluation scenarios and feedback loops
• Analyse outputs to identify failure modes and system weaknesses
• Design for reliability at the system level, preventing agentic breakdowns
• Shape direction from the earliest stages of zero-to-one problems
• Influence across teams without formal authority
• Operate in a highly autonomous, fast-moving environment with evolving constraints
• Lead ambiguous, in-flight work—making calls on when to adapt the product vs. push on system improvements as things evolve
• Act as a bridge between AI research, technical capabilities, and user needs
• Collaborate deeply with ML scientists, PMs, and product designers
• Prototype and run focused experiments that isolate variables and generate clear insights
• Stay close to system behaviour by regularly reviewing outputs and raising the bar on quality