Principal Product Designer, Shield AI

• 12+ years designing and building production software interfaces, with end-to-end ownership from research through shipped product
• Expert-level React/TypeScript; you write production-quality front-end code, not just prototypes
• Portfolio of shipped tools or applications demonstrating strong interaction design and visual craft
• Hands-on experience designing LLM-powered experiences: streaming, progressive rendering, loading states, context and conversation management
• Deep knowledge of client and UI state management, the DOM, browser APIs, WebSockets, and Server-Sent Events
• Experience running user research and usability testing, and acting on the results
• Strong software hygiene: Git, code review, testing, CI/CD
• Fluid intelligence that allows one to operate effectively in ambiguous conditions while finding opportunities to drive technical efforts and force multiply
• Experience with Linux, Docker, and a major cloud platform (Azure, GCP, AWS)
• This role requires travel to a Shield AI location 1-3 times per quarter

Desirable

• The Special Projects team at Shield AI is an elite force within the office of the CTO
• It consists of a group of very senior (L5-L8) and highly experienced software engineering experts from diverse fields (aerospace, robotics, cloud infrastructure, game development, interactive media design, ...)
• The charter of this group is to steer technology development towards strategic alignment with the CTO's vision, through tactical insertion into teams and technologies across the organization
• Individuals within this team make direct and at times forward-sprinting contributions to all three pillars of Hivemind, Shield AI's software ecosystem for developing and deploying resilient intelligent teaming for aircrafts
• Hivemind consists of four products: EdgeOS (C++ based high-performance middleware for autonomy development), Pilot (autonomy for the edge built atop EdgeOS; a models-based, modular and open architecture C++ codebase), Forge (Shield AI's AI Factory for the design, development, and testing of Hivemind Edge systems; a service-oriented architecture leveraged through an SDK, CLI, and web portal; a Go, Python, Typescript codebase), Commander (Software and hardware to support rich human-in-the-loop and human-on-the-loop interactions with the Hivemind; a C++-based back-end for interaction with Pilot; and a web-application UI for mission planning, command and control by operators implemented in a Typescript/React codebase)
• The Special Projects team is chartered to operate effectively within ambiguity
• While this team does not own any software products, it tactically and strategically impacts the development of all foundational Hivemind products and leverages those technologies to rapidly deliver low-TRL solutions
• This work happens shoulder-to-shoulder with the product teams in some cases, and in a forward-sprinting manner within the Special Projects team in other cases
• The result is direct contribution to products in the former case, and development of reference implementations in the latter
• The Special Projects team also functions as a pipeline for product and solution engineering teams across the organization
• Individuals who enter the Special Projects team rapidly gain depth and breadth in their understanding of the Hivemind software ecosystem
• This positions them well for leading technology development effort across the product and solution organization
• Design and build the user experience for agentic AI products across the office of the CTO: an enterprise knowledge assistant (Brain), an agent-driven engineering platform (Springboard), and enterprise AI/data applications
• Prototype in code. Ship working React/TypeScript implementations, not mockups — then land them in production with the product teams
• Design the core interaction patterns of agentic software: streaming responses, citations and provenance, multi-modal input, artifact sharing, and human-in-the-loop review and approval surfaces
• Design interfaces for foundation-model based autonomy architectures and model lifecycle tooling: evaluation dashboards, trace inspection, observability, and verification surfaces
• Conduct user research across engineers, operators, and executives; define personae and translate findings into design decisions
• Investigate competing agentic and enterprise AI products; synthesize findings for engineering and product leadership
• Establish reusable design patterns and a component vocabulary for AI applications across the organization
• Drive adoption: instrument what you ship, watch real usage, and iterate daily