Senior Product Engineer
SageOx
Senior Product Engineer
SageOx
Shared memory for human–AI coding teams
SageOx is a well-funded, early-stage startup at the frontier of how AI is shaping the world. At SageOx, we believe the future of work is deeply human and collaborative. As agents take on execution, coordination, and routine work, people gain space for what only humans do well: creativity, judgment, relationships, the work of imagining, deciding, and creating together. Humans become not only more productive individually, but better at working with each other, leaning on their AI coworkers to carry out execution and raise the quality of how they collaborate.
We're looking for strategic hires - fearless, creative, independent thinkers - that will shape the future of the company, and how work is done in the modern world.
There's no trail yet. Pioneering and breaking trails is one of our company's core values. You're one of the first ten people cutting it.
Everyone touches all layers. That's not a slogan. It's the actual job. Full-stack at SageOx really means the whole stack: hardware firmware sitting in the room, mobile/desktop apps, cloud services, Temporal workflows, web frontend, real-time streaming, CLI, and file systems.
At the speed at which we're moving, we believe being in person and collaborating closely is essential. This is an in-person, Seattle-based role. We strongly believe that magic happens when people are in the flow together working on ambitious problems.
Pick a week at random and you might be:
Researching and building out episodic and semantic memory systems that deviate from the standard mental models everyone else is chasing.
Chasing a lock-free ring-buffer bug in the Ox Dot hardware firmware that silently caps every recording at twenty minutes.
Hardening a Temporal workflow that's processing recordings at 2am so it survives a pod eviction without losing a byte of audio. Recordings don't get a second take.
Fixing mobile audio pipelines so flaky hotel WiFi can't quietly drop a recording that nobody notices is gone until the meeting's long over.
Exploring real-time audio processing and speaker identification improvements to ensure every captured nuance is correctly attributed for agents to really know the team.
Writing native macOS/Windows code that lets a desktop app capture both sides of a call with no virtual audio driver.
Shipping a new feature defined yesterday into the web product that a user starts working with minutes later and never thinks about the complexity underneath.
If "that's not my code" is a sentence you reach for often, the day-to-day here will grate. If you don't understand why anyone would say it, you'll probably feel at home.
We have some peculiar tendencies, which make our work far more impactful, fast, and joyful, but they are not for everyone. Radical transparency isn't a value on a poster: it's how we work. We record the entire engineering journey: every coding session, every AI conversation, every wrong assumption, every rabbit hole, every breakthrough. We don't hide unfinished work because that's where the learning lives.
Roughly: Go/Rust/Swift, TypeScript/Next.js, React Native/Expo/Electron, Postgres, Temporal, AWS/K8s, embedded C/C++, MoQ. You won't touch all of it your first week, but nothing here is permanently someone else's problem.
We ship fast, but speed was never the actual bar. Taste is. That's the bar, both halves at once: architectural taste (reach for the fix that makes the bug structurally impossible, not the patch) and product taste (care about the four seconds nobody will ever thank you for).
We're not optimizing for a particular pedigree. We're looking for entrepreneurial builders with systems-first thinking, AI chops, and product judgment. The rest is just evidence.
AI-pilled, for real. Agentic coding — Claude Code, Codex, OpenCode, whichever tool — is already your daily practice, not something you tried once. We won't teach you this on the job; the work trial assumes you show up fluent.
You write harnesses, loops, and specs — not just code. Hand-crafting every line was the whole craft for years. It isn't anymore. The leverage is in the system that produces the code — the loop, the guardrails, the verification — while you bring the judgment. If watching an agent write your code makes you anxious instead of curious, we're not a match.
Tinkerer and explorer. Side projects, weekends spent building something nobody asked for, a GitHub full of half-finished ideas. You can't stop dreaming up the next thing and poking at how it might work. We don't care whether any of it succeeded, we care that you couldn't stop trying.
Curious sideways, not just deep. You've probably gone down a rabbit hole on something outside your lane. How a database index actually works, why a mic distorts at certain frequencies, what makes an animation feel expensive. You do this because you wanted to know, not because anyone assigned it. Borrowing an idea from one domain and slamming it into another is most of what "full-stack" means here.
7+ years building and shipping production software. This is a senior role. You’ve owned systems end to end, lived with your own decisions in production, and know which corners are safe to cut — and which ones come back at 2am.
A track record you can point to. Doesn't have to be famous. Has to be real.
You finish, and you ship. A hard problem is fun to poke at; the job is turning that into something running in production. Not a forever-branch, not a slide deck nobody acted on. If shipping is the part you rush past, we're not a match.
You take the unproven path on purpose, sometimes. Calculated risk, not recklessness — a pulse of someone who'd rather build the new thing than maintain the safe one.
Bonus, not required: you've led before. A surprising number of the people who thrive as IC engineers in an agentic era managed teams at some point — engineering managers, tech leads running multiple workstreams, people who've had to break an ambiguous problem into pieces and hand them to other people. That's not a coincidence. Directing a fleet of AI coworkers through a hard problem draws on the same muscle as directing a team of humans through one: decompose clearly, delegate with real intent, review what comes back, know exactly when to step in and rewrite it yourself.
None of this is about pedigree, and we're bad at predicting it from a resume anyway — that's the entire thesis of the post linked below. If you read this list and land on "probably," the work trial is a faster, more honest way to find out than we could ever manage in an interview loop. This interview goes both ways. You learn as much about us, as we do you, and whether you gel with our style of working together. The trial gives both parties the most honest data about that fit.
We don't run technical interviews. No whiteboard algorithms, no four-hour system-design rituals. Clear an initial assessment and we'll invite you to a work trial: 3–5 days, full-time, in person, building something real alongside the team, the way the team actually works. The trial will be on terms that are mutually agreed upon. Most of what either side needs to know about fit shows up in those few days, not in a loop of hypotheticals.
Ajit, our CEO, wrote up why we replaced interviews with this, including the uncomfortable math behind it.
$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. Staff/Principal candidates may get basic relocation help.
There's no shortage of comfortable places to write software right now. This is not one of them, and that's the point.
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