Design Engineer
SageOx
Design Engineer
SageOx
Shared memory for human–AI coding teams
You'll be the second designer at SageOx, partnering closely with our Founding Designer — not joining a design org, not reporting into one. Before you apply, read her post about what this actually feels like. It's the most honest preview of this role that exists, including the uncomfortable parts.
We're not looking for another generalist. We want a T-shaped designer: exceptional at one thing — good enough to take part of our product somewhere none of us could have imagined — and strong enough to play across everything else. In the AI era, someone who is a rock star in one area can push out and have real influence in the others.
SageOx doesn't live on one screen. It lives in the product people use every day, on the conference room screen a whole team looks at together, and in hardware that sits on the table. You'll design and build across all of it, along with the agentic design harness that lets every AI coworker on the team build the way you would.
And you'll push us. We want someone who tells us we aren't thinking ambitiously enough about the experience — then ships the version that proves it.
These are shipping surfaces, not concepts on a roadmap, and most of them have no obvious precedent.
Design features end to end, from your own user research to the shipped experience — web, desktop, mobile, CLI. Designers here write the production code. If you can't build it, the design doesn't ship.
Memory you can walk up to. What a team has said, decided, and built. Designing retrieval people trust — where a claim came from, what an AI coworker actually read — is an information-design problem worthy of Tufte.
Ambient coordination. Murmurs — small signals from your coworkers, human and AI, that reach you without becoming another notification you resent.
How people grow into SageOx. We're breaking trail on human–AI collaboration, so the learning curve is part of the product: how people start, what nudges them next, and how they get more value over time.
Talk to users, and use SageOx daily. Everyone here does. Design from lived use.
A room, cast to a screen. A live conversation on the TV everyone is already looking at, legible from across the room. Different rules than anything designed for arm's length.
A living mural. A meeting draws itself as it happens, in front of everyone in the room. What should a thought look like thirty seconds after someone says it? Nobody knows yet.
The room as a surface. What a meeting room becomes when a team's memory and its AI coworkers are present in it, and how everyone in it can see and steer what's happening.
A device with almost no screen. The Ox Dot sits on the table and gives the hivemind a body in the room. Its whole vocabulary is presence, light, and sound. Every instinct you've built for a rectangle is the wrong instinct here.
SageOx in the real world. Packaging, unboxing, first run, and pairing — every moment someone holds something of ours in their hands.
What comes next. A watch, new devices, and whatever we decide to build after that. Work out what our patterns become on each one, and lead those builds.
Own the design system. Components, tokens, patterns, and states across every surface above, built in code, with accessibility built in so every feature inherits it.
Make it readable by AI coworkers. Docs, skills, and rules that Claude Code reaches for before it invents something new. When an AI coworker gets it wrong, treat that as a gap in the harness and close it. Our litmus test: can a future AI coworker read the UI's state, change it correctly, and confirm the result with no human in the loop?
Catch drift before users do. Turn design rules into automated checks. Sample what shipped this week — at random, not curated — find the drift, codify the rule, automate it, and verify it stopped. The goal is a team that ships good design without a design approval gate.
Turn the harness into product. Parts of it become agentic tools we ship to customers, starting with design walkthrough tooling in our desktop app.
This job is done in person, in Seattle, five days a week. Not remote. Not hybrid. Not "in person to start." At the speed we're moving, being in the same room is essential, and we strongly believe that magic happens when people are in the flow together working on ambitious problems. If you're looking for a remote or remote-friendly role, please don't apply to this one — we'd rather you not spend the time.
Two narrow situations we'll still talk to you about. Both are in person; neither is remote work:
You'd be relocating to Seattle. For select Staff/Principal-level candidates, we may help with some basic relocation costs. You'd be here, in the office, like everyone else.
You're in the Bay Area. Through the end of 2026, we may consider a San Francisco–based candidate who can be in Seattle, in person, at least two weeks a month.
That's the whole list. Outside of those two, this role requires living in the Seattle area and working alongside the team on site.
We're not optimizing for pedigree. We're looking for builders with design judgment — and we want to see what you've built.
T-shaped, with a real spike. You're exceptional at one thing — product UX and interaction, hardware and physical product, large-screen and ambient interfaces, or design systems and tooling — and solid across the rest. Show us the spike.
Strong, opinionated taste. You'll tell us when we aren't thinking big enough about the experience, and back the opinion with something you shipped.
A tinkerer who builds things, not just designs them. Apps, tools, side projects, a home-lab rig, a device you hacked together on your bench — things you made yourself, end to end. We don't care whether any of it succeeded; we care that you couldn't stop building.
5+ years in product design, design engineering, industrial design, design systems, or front-end engineering. Long enough to know the design rules, and when to break them. The title matters less than the evidence: what you built, and what happened to the product after.
AI-pilled, for real. Claude Code, Claude Design, or your own stack is already your daily practice. You've built something AI coworkers use — a skill, an MCP server, a lint rule, a generator. Show us.
Technically fluent, with craft. At home in a codebase and in the terminal. You notice the empty state, the focus ring, and the error message. No CS degree required.
Working software over polished mockups. You balance speed with taste. In an age of AI slop, that often means editing and systematizing rather than producing more.
Curious sideways. Why a mic distorts, how an index works, what makes an animation feel expensive.
Ownership, low ego. You own outcomes, make things happen with limited resources, and find direction in early-stage ambiguity. Strong opinions in crit, open to new perspectives.
You believe in what we're building. A world where people and AI coworkers work as one team — and you want to be the one designing it.
Hardware, industrial, or packaging design experience on a product that shipped.
Interfaces for large shared screens: TVs, conference rooms, signage.
An engineering background, or time at a developer-tools company.
Growth or onboarding experience — designing education and activation.
Founder or early-employee experience somewhere that grew fast.
The best part of your week is a pixel-perfect spec that someone else implements.
Your portfolio is all mockups and case studies, with nothing you built yourself.
You'd rather go wide on everything than go deep on anything.
Watching an AI coworker write your code makes you anxious rather than curious.
Design review, to you, is a gate you hold. Here it never blocks a ship — we ship, then reconcile with a loop, so no human sits in the critical path.
None of that is a knock on great designers. It's just not this job.
No whiteboard interviews, no take-home redesign of somebody's app. We start with your portfolio. Clear an initial assessment and we'll invite you to a work trial: 3–5 days, full-time, in person in Seattle, building something real alongside the team, the way the team actually works. It's paid, and the terms are agreed with you up front. Most of what either side needs to know about fit shows up in those few days, and it goes both ways — you learn as much about us as we do about you.
Before applying, please learn more about our work trials.
Ajit, our CEO, wrote up why we replaced interviews with this, including the uncomfortable math behind it: Read the post →
$15M seed led by Canaan, with A.Capital, Pioneer Square Labs, and Founders' Co-op. Founded by Ajit Banerjee (AWS EC2/EBS, XetHub → Hugging Face), Ryan Snodgrass (early Amazon, Kindle/AWS-scale cloud systems), and Milkana Brace (EVP Product at Remitly) — people who've built this kind of infrastructure, and scaled through an acquisition or two, before. We're small on purpose, hiring our first ten. More about the team →
Competitive salary and meaningful equity. We'll talk specifics once we know it's a real mutual fit. That's what the trial is for.
There's no shortage of comfortable places to design software right now. This is not one of them, and that's the point.
Include your portfolio, and lead with your spike. Show us the one area where you're exceptional, then the things you built yourself — an app, a tool, a device, a live site, a skill or MCP server — next to the design work behind them. For each, tell us what you made, how you built it, and what happened after. A repo or a live link beats a polished case study.
We're not working with recruiters or recruiting agencies/consultants at this time — direct applications only, please.
SageOx builds agentic context infrastructure to keep AI coding agents and human teams aligned by creating a shared memory of decisions, constraints, and reasoning from discussions and agent sessions. It operates through four components: Team Context to define governing principles, Ledger of Work to record reasoning and intent, the Ox CLI to prime agent sessions with shared context, and a web app to search and navigate the history, linking code commits to original intent. The platform differentiates itself by enabling multiplayer by default, so every human and AI agent draws from a versioned, common context instead of working in isolated sessions, which helps prevent architectural drift. Its goal is to prevent context loss and rework in AI-assisted software development, delivering a SaaS product for AI-native teams (initially focusing on Claude/Anthropic models) to support aligned human-agent collaboration throughout the software development lifecycle.
Explore more opportunities at SageOx