Staff Product Designer, Agentic Coding

RDQ427R613

About Databricks

Databricks is the Data and AI company. More than 10,000 organizations worldwide rely on the Databricks Data Intelligence Platform to unify and democratize data, analytics, and AI. As agents move from demos to real production work, Databricks is building the platform that lets enterprises create, govern, and collaborate with coding agents safely and at scale.

Team & Product

Omnigent is Databricks' platform for building, composing, governing, and sharing AI coding agents. It lets people create and coordinate agents across different model providers and execution environments - from the CLI and IDE to desktop, mobile, and managed enterprise surfaces. Users can watch agents work live, jump in to co-drive or fork a session, and package and share reusable agents.

This is a rare chance to define and refine interaction patterns at enterprise scale: how people supervise fleets of agents, intervene at the right moment, understand risk, and build trust in autonomous work - starting with coding, review, and exploration workflows. The hard design problem isn't just showing what agents are doing; it's helping people know when to trust, when to step in, and how to recover when work goes off track - across a high-density desktop workspace and lightweight mobile moments.

You'll design for two audiences at once: an open-source community shaping the future of agent workflows, and enterprise teams adopting agents in governed production.

Role

You'll be the IC design lead for how humans and fleets of agents work together on Omnigent - leading through product strategy, systems thinking, craft, and cross-functional collaboration. You're equally comfortable shipping small design changes directly in the codebase and prototyping your way through big, ambiguous problems that don't have a spec yet.

You'll own the design direction across desktop and mobile, partner deeply with engineering and product, and translate the frontier of what agents can do into experiences that feel clear, controllable, trustworthy, and surprisingly effortless.

What you'll do

• Own the desktop and mobile experience for live agent sessions end to end - from strategy and interaction models to visual polish and shipped detail. Shape the high-density orchestration workflows on desktop and the lightweight monitoring, review, approval, and intervention moments on mobile.
• Design directly in the codebase for well-scoped changes, working alongside engineers to ship real improvements rather than handing off static mocks.
• Use prototypes as a primary thinking tool. For large, vague problems, build interactive prototypes to explore, de-risk, and align the team on a direction.
• Invent the interaction language for agentic work: real-time multi-agent sessions, watch/co-drive/fork flows, human-in-the-loop approvals, agent status and transparency, error and recovery states, and trust.
• Define durable UX principles and interaction frameworks for agent collaboration, approvals, observability, and recovery that other teams reuse across Omnigent surfaces.
• Drive cross-platform consistency with a strong, scalable design system spanning desktop and mobile.
• Ground decisions in evidence - use research instincts and real usage to pressure-test ideas and know what to build next.

What we're looking for

• A builder-designer who works fluently with AI tools. You can make small design changes directly in a codebase, and you use prototypes - and the agents themselves - to explore larger, ambiguous problems. You regularly use AI tools in your own design and build workflow.
• A portfolio of strong end-to-end craft - interaction design, visual/UI polish, and design-systems thinking - ideally on developer tools, AI/ML products, enterprise software, or other products for technical users.
• Staff-level scope and impact. A track record of shipping complex, high-quality developer products and leading design direction on hard problems.
• Self-directed, and comfortable in ambiguity - you help shape product direction when the path isn't defined.
• A strong collaborator who raises the level of high-quality critique, clear decision-making, and tight partnership with engineering and product.

Nice to have

• Experience designing coding tools, IDEs, agent frameworks, or developer platforms.
• Familiarity with multi-agent systems, LLM harnesses, or governance/observability tooling.