About this roleWe're hiring a seasoned Designer to improve how design and engineering work together, starting with a world-class, agent-first design system.
By agent-first, we mean the system is structured so humans and AI agents can make the same good decisions: choosing the right components, applying the right patterns, using consistent names, and reviewing UI against shared rules.
This is a systems thinker, high-craft, high-collaboration role. You'll work closely with dedicated engineers who will be your collaborators in bringing this to life across our product surfaces. You'll own the design quality, semantics, and product usefulness of the system; engineering will own the technical architecture and implementation, with close collaboration on component APIs, constraints, and adoption.
Scope- You'll work across iOS, Android, and web.
- We're mobile-first in the near term, with an expectation that components and patterns translate cleanly to web as those surfaces expand.
- This role focuses on product surfaces (not the marketing site).
What you'll own- Design ↔ Engineering workflow improvements. Treat the seams between functions as a design problem: shared language, fewer re-made decisions, clearer handoffs, and faster shipping.
- Component craft and system coherence. Set the quality bar for components and patterns, and pressure-test them by building and feeling the real thing (ideally prototyping with agents), not only speccing.
- The semantic layer that scales. Component naming, usage rules, and decision guidance so the right choice is the easy choice for any designer, engineer, or agent.
- Adoption across the design org. Create the standards and workflows that make using the design system the default, not something teams and agents have to seek out.
- Documentation + QA workflows. Own usage documentation and quality check workflows that make quality as hands-off and reliable as possible, including AI-assisted checks for consistency, accessibility, and component usage.
What you'll build firstIn your first months, you'll likely:
- Build and evolve the components we actually need - prioritizing the highest-leverage gaps, pressure-testing them in real product work, and turning repeated decisions into durable patterns.
- Establish agent-first systems and workflows - rules + semantics that help humans and agents use the system correctly without depending on tribal knowledge. Set them up to show up in the right moments to create efficiency and consistency across teams.
- Set the operating cadence with design and engineering - a contribution/review rhythm that keeps the system coherent and adopted.
What we're looking for- 6+ years in product design, with real depth in design systems built on tokens, component APIs, and a single source of truth in code (not only a Figma library).
- You've helped build a design system before. You've contributed meaningfully to components/patterns/tokens, governance, adoption, and the messy realities of getting it used.
- A systems thinker. You'd rather build the thing ten people use than ship one feature yourself.
- Taste you can defend. You can pinpoint what's off (spacing, motion, interaction states) and explain what's right for Speak and why.
- Clear communication + cross-functional drive. You can write the rule, not just draw the component, and work across teams to drive adoption.
- Autonomy and end-to-end ownership. You thrive with little scaffolding and drive from problem → roadmap → shipped → adopted.
- Fluency with AI in your craft. You've built repeatable ways of working with AI/agents and can explain what changed and why.
- Accessibility and scale mindset. You think about how the system holds up as it grows across mobile and web.
Nice to have- Comfort building or changing components in code (Swift/SwiftUI, Kotlin/Jetpack Compose, or web).
- Shipped consumer-scale mobile products across many markets/locales.
- Built a design system engineers actually used - and that outlasted your time on the team.
- A point of view on where design-to-code and agent workflows are heading, and what's still broken.
- Experience in language learning, education, or another domain where cross-cultural nuance shapes the product.
