Senior Design Engineer

AION•San Francisco, CA

12h

About The Position

aion is the enterprise AI platform, a full-stack solution for building, fine-tuning, and deploying AI at scale. Whether an organization is modernizing internal operations, launching AI-powered products, or transforming customer experiences, aion takes them from concept to production on a single, unified platform. We’re a fast-growing, VC-backed startup led by founders with a track record of successful exits. With teams across the US, UK, and India, we’re building the next generation of enterprise AI and we’re looking for exceptional people to help us scale. We're hiring a design engineer who builds like a founder. You see what's missing and make it exist: you design it, build it, ship it and own it. Code is your design material, and you write a shader when CSS runs out. The bar is an interface that feels physical: it tracks the finger, carries momentum and never drops a frame.

Requirements

5+ years shipping production front ends, with live work you can link to.
React and TypeScript at expert level, with component APIs that don't lie: contextual props, derived state, variants over boolean flags and compound components.
Motion values and direct DOM updates for anything high-frequency, so gestures and scroll never wait on a render.
CSS in depth: clip-path, masks, backdrop filters, blend modes, scroll-driven animations, container queries and the View Transitions API.
Type and pixels in code: variable fonts, optical sizing, OpenType features, and layouts free of shift and subpixel blur at 1×, 2× and 3×.
Performance as a feature: an 8 ms frame budget at 120 Hz, compositor-only animation, and pages that load instantly because every byte was questioned.
Accessibility as craft: focus rings styled rather than removed, focus moved into overlays and back, full keyboard paths and reduced-motion settings honoured.
Real-time 3D with Three.js or React Three Fiber on WebGL or WebGPU.
Shaders in GLSL or WGSL: noise, signed distance fields, glass, refraction, lighting and post-processing, all inside the frame budget.
Motion studies in After Effects and 3D in Cinema 4D or Blender, so ideas move before a line of production code exists.
Generative and data-driven work in Canvas and SVG, and interactive animation in Rive.

Nice To Haves

Custom tools you built that others rely on: plugins, design and motion tooling, debug overlays, generators or internal apps.
Something you started and shipped yourself: a product, a studio, a library or a side project with real users/customers.
Writing and articles about your craft: essays, breakdowns or notes that show how you think.
Creative coding or generative art.
Open-source work in motion, graphics or UI components.

Responsibilities

Take ideas from a blank file to production: decide what's worth building, prototype it in code, and own it once it ships.
Build production UI in React and TypeScript, from primitives to complete surfaces.
Build the custom tools that make the team faster: design-system tooling, motion tuners, debug overlays and internal apps.
Tune motion and gestures until they feel physical: momentum, weight and interruption.
Create 3D, shader and generative work where it lifts the experience.
Review every build on real devices at 1× and 2×, and keep it fast, accessible and robust under abuse.
You treat interaction design as thinning the layer between intent and software, and your work has moments where the interface inferred what someone wanted instead of asking.
Physics is a toolkit to you, not a style: springs tuned by stiffness, damping and mass; velocity projected on release so a flick finishes the job; rubber banding where a hard stop would feel broken; secondary elements interpolated from the gesture driving them.
You give movement conceptual weight: light things settle quickly, heavy and destructive things resist, and bounce appears only where something was thrown.
Your gestures have three beats, constrain on start, track 1:1 on move, resolve on end, and stay contained: no stray text selection, no triggered neighbours, no pointer lost at the edge.
Choreography is second nature: the primary element answers instantly, secondary elements follow late, exits clear faster than entrances arrive, and a touch of blur hides the seams.
Your restraint scales with frequency and input: touch gets the most feedback, the pointer some, the keyboard almost none, and anything used a hundred times a day loses its animation.
You build without waiting: when a tool, component or prototype doesn't exist, you make it, and you've shipped things of your own that people use.
Code is your design material. You judge an idea by building and using it, prototype in the browser rather than in static frames, and polish only once the idea has earned it.
You iterate until nothing is left to explore, then have the discipline to choose the simple version from many iterations back.
Robustness is binary to you: you spam every state, scroll too fast, drag past every edge, and cut any detail that can't work every time.
You see at the pixel: optical alignment at high zoom, animations scrubbed frame by frame, and before-and-after captures on every change.
Native first: real scroll over hijacked wheel events, clip-path over animating width, CSS keyframes before reaching for a library.
You study the work you admire by rebuilding it, then make something unmistakably your own.
Interfaces for streamed, non-deterministic output: progressive rendering, interruption and retry.
You build daily with AI coding tools such as Claude Code, Cursor or Codex, use them to ship whole tools and products on your own, and review every line they write.
Work directly with high-pedigree founders shaping technical and product strategy.
Build infrastructure powering the future of AI compute globally.