Staff Product Designer Design System positions focus on delivering results in their domain. This page aggregates open Staff Product Designer Design System roles and what employers typically expect.
About the Team People love using HubSpot because it’s an essential product that gives them the tools and insights they need to grow their business. Trellis is HubSpot’s enterprise design system, the shared foundation that powers product surfaces across CRM, Marketing Hub, Sales Hub, Service Hub, Commerce Hub, Breeze AI, and the growing portfolio of agentic experiences. Trellis is not a style guide or a Figma library. It is the connective tissue between design intent and product reality at scale: a living system of tokens, components, patterns, documentation, and tooling that designers and engineers across HubSpot depend on every day to build with speed, quality, accessibility, and consistency. We are at a defining inflection point. AI is fundamentally expanding what a design system needs to be. In the AI native era, Trellis must be machine readable as well as human readable. Tokens need semantic structure that AI tools can reason about. Components and patterns need usage guidance that helps teams make better decisions earlier. Documentation needs to serve the designer working in Figma, the engineer implementing in code, and the AI assisted tool generating product UI. This is the environment you’re stepping into, and it is exactly the right moment to shape how HubSpot builds. About this Role We are hiring three Staff Product Designers to help shape the next phase of Trellis, HubSpot’s enterprise design system. These roles share the same broader mission: helping product teams across HubSpot build faster, more consistently, and with a higher quality bar as our platform becomes more connected, adaptive, and AI native. Each role will focus on a different part of the system: Tokens and Visual Language This role owns the foundational visual system behind Trellis, including color, typography, spacing, iconography, motion, theming, density, semantic tokens, and the visual quality bar for how HubSpot shows up in product. This person will help evolve Trellis’ token architecture and visual language so it can scale across product surfaces, support AI assisted development, and remain coherent as the platform grows. Components This role owns the reusable building blocks product teams use every day, including component anatomy, behavior, variants, states, accessibility, documentation, implementation guidance, and Figma to code alignment. This person will help define what belongs in the shared system, how components should behave, and how Trellis can reduce fragmentation across teams without slowing product development. Patterns This role owns the repeatable product behaviors and interaction models that show up across HubSpot, such as creation, editing, filtering, search, object management, permissions, notifications, recommendations, and AI assisted workflows. This person will help product teams solve common problems in more consistent ways and make HubSpot feel more coherent across connected customer workflows. Across all three roles, you’ll work at the intersection of product design, systems thinking, craft, engineering partnership, governance, adoption, documentation, and implementation quality. You will make decisions that affect many product teams and help define how a large B2B platform builds product in an AI native era. This is not a role where you simply maintain what exists. Trellis is in a multi-year modernization journey, and this team is helping make the system trusted, self-serve, and useful in day-to-day product development. You will help reduce fragmentation, improve system adoption, shape contribution models, and create the shared foundations that let HubSpot build with more speed and quality at scale. In this role, you’ll: Own and evolve a major focus area within Trellis, either Tokens and Visual Language, Components, or Patterns Partner with product, engineering, content design, research, and design technology to define system guidance that is usable, scalable, accessible, and practical to implement Make high-judgment syst…