Staff AI Designer

Fin is the AI Customer Agent company on a mission to help businesses provide perfect customer experiences. Our AI Agent Fin is the highest-performing AI Customer Agent on the market today, enabling businesses to deliver impeccable, always-on customer support across the customer journey – from service, to sales, to ecommerce. Powered by our own AI models, Fin resolves complex customer issues end-to-end across every channel, with minimal set-up and integration. Fin can also be combined with our natively integrated Intercom help desk for one single system that is designed to meet the needs of modern day support teams. Founded in 2011, Fin became one of the fastest growing companies and remains one of the largest private software companies in the world with nearly 30,000 global businesses using our products to transform their customer support. Driven by our core values, we push boundaries, build with speed and intensity, and relentlessly deliver incredible value to our customers. Role Overview We’re looking for a Staff AI Product Designer to shape what our AI does, how it behaves, and how it fits into real user workflows. This role sits at the intersection of product, design, and machine learning—turning emerging capabilities into reliable, high-quality user experiences. You’ll design systems: behaviors, decisions, orchestration, and end-to-end experiences powered by AI. This is not a traditional product design role. You’ll go beyond interface design to define how AI systems behave, operate, and deliver value in the real world—similar to emerging “AI Model Designer” roles, but more deeply embedded in product. What You’ll Do 1. 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 2. 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 3. 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 4. 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 - Analyze outputs to identify failure modes and system weaknesses - Design for reliability at the system level, preventing agentic breakdowns How You’ll Work - 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 behavior by regularly reviewing outputs and raising the bar on quality What We’re Looking For - Strong product judgment—you consistently make high-quality calls on what to build and why - Ability to influence without authority - Experience working on zero-to-one products and comfort operating through ambiguity and rapid change - Deep understanding of LLMs and their limitations, along with a grounding in traditional ML approaches and when to use them instead - Experience defining design principles, heuristics, or evaluation criteria for AI systems - High standards for quality, with a track record of identifying where AI systems break down in real-world use - A scientific mindset—you can design experiments that isolate variables and produce meaningful insights - Systems thinker: you’re happier designing behaviors and flows than polishing interfaces - We’re open to non-traditional backgrounds (e.g. product managers, engineers, or others) if you have strong product sense and a tra

Strong product judgment; ability to influence without authority; experience working on zero-to-one products and comfort operating through ambiguity and rapid change; deep understanding of LLMs and their limitations, along with a grounding in traditional ML approaches and when to use them instead; experience defining design principles, heuristics, or evaluation criteria for AI systems; high standards for quality, with a track record of identifying where AI systems break down in real-world use; a scientific mindset with the ability to design experiments that isolate variables and produce meaningful insights; systems thinker who prefers designing behaviors and flows to polishing interfaces; openness to non-traditional backgrounds (e.g., product managers, engineers) if you have strong product sense.

Define AI capabilities (the “what and why”): identify high-value opportunities for AI, define what AI should do and what it should not do, set boundaries for autonomy vs human control. Design AI behavior, orchestration, and system logic: architect AI-driven workflows and agentic systems, establish decision frameworks, design for uncertainty and edge cases, develop design principles to shape AI behavior, ensure coherence as complexity scales. Shape human + AI workflows and end-to-end experiences: design experiences for users managing AI systems, integrate AI into real workflows, define the role of humans in the loop. Define quality, evaluation, and reliability: specify what “good” looks like, design evaluation scenarios and feedback loops, analyze outputs to identify failure modes, design for reliability at system level.