Principal Agent Designer

Principal AX (Agentic Experience) Designer

As a Principal AX Designer at Out Systems, you pioneer a new design discipline where the “user” extends beyond humans to include autonomous AI agents. You will define how agents navigate, operate, and succeed within our platform while preserving clear, trustworthy human oversight when needed. This role sits at the frontier of UX → UAX → AX evolution and involves collaborating with Product Management, Engineering, AI teams, and Research to transform the platform into an agent‑ready substrate that is machine‑legible, safe by design, observable by default, and cost‑aware.

• Design the Agent Interface Layer.
• Define and design the machine‑operable substrate beneath and alongside the UI.
• Make inter‑agent and agent‑to‑internal and external systems communication clear, intuitive, and agent‑readable.
• Define and develop new design assets required for AX, including experience artifacts, specifications, and governance frameworks for agent operation (e.g., operation contracts, behaviour policies, trust specifications, evaluation criteria).
• Lead the AX design lifecycle:
  Model → Contract → Constrain → Observe → Validate → Run → Learn → Version.
• Define operability gates for every agent workflow: determinism, safety, observability, recoverability, and economics.
• Build evaluation frameworks such as golden tasks, regression sets, and adversarial scenarios to gate capability version changes.
• Design for all three oversight modes (interactive, supervised, lights‑out), specifying the appropriate mode per operation and risk profile.
• Bridge UX, UAX, and AX design dual‑layer journey maps with human experience and agent operations as parallel swim lanes, identifying delegation moments and oversight touchpoints.
• Define an operating model for trust per operation: when humans interact, when they supervise, and when agents run autonomously.
• Champion dual legibility: ensure every element is readable by humans and interpretable by agents.
• Shape strategy and practice: influence product vision and department strategy for agent‑readiness and AI‑native design.
• Establish AX design standards, patterns, and best practices across the R&D organization.
• Contribute to the evolution of the design system to support agent‑facing components, semantic structures, and observability surfaces.
• Evangelize agent‑first design thinking across Product, Engineering, and leadership.
• Lead and develop people: mentor a team of designers through the UX‑to‑AX transition, building fluency in semantic architecture, process modelling, and exception design.
• Drive reskilling from empathy‑first to precision‑first design: move from user research to process analysis, from emotional design to semantic architecture, and from session thinking to system optimisation.
• Foster collaboration and a shared vision for design excellence across all three experience eras.

• 8+ years of experience in UX, Product Design, or related design disciplines.
• Demonstrated experience designing complex B2B SaaS or platform products.
• Strong foundation in systems thinking, information architecture, and semantic design.
• Experience building operating models and conceptual design.
• Experience leading cross‑functional teams (design, engineering, product, AI/ML).
• Experience with the full design lifecycle: research, ideation, prototyping, testing, iteration, and delivery in Agile environments.
• Expertise in Figma and other modern design tools (Figma‑make, Cursor, Claude Code, Codex).
• Excellent communication skills and ability to articulate complex agent‑system interactions to technical and non‑technical stakeholders.
• Fluency in English, both spoken and written.
• Understanding or hands‑on experience with Prompt Engineering in software development.
• Familiarity with AI‑native IDEs or assistants such as Cursor, Git Hub Copilot, and Claude.
• Deep fluency in LLM patterns (ReAct, Chain‑of‑Thought) and ability to design state‑machines, tool‑use contracts, and autonomous workflows for agents.
• Experience designing for non‑deterministic systems, including graceful failure handling and recovery paths.
• Mastery of semantic…