Skip to main content

AI Engineer - Backend

Posted yesterday

agiMillbrae (CA)

SENIORITY

Lead

Apply (opens in a new tab)

About the role

Think Different. Build the Future. Build everyday AGI. Trustworthy, consumer-grade agents that redefine human–AI collaboration for millions. Software shouldn't wait for commands; it should partner with you, amplifying what you can do every single day. Why AGI, Inc. We're a stealth team of elite founders and AI researchers, with backgrounds spanning Stanford, OpenAI, and DeepMind. We're industry leaders in mobile and computer-use agents, bringing these capabilities to consumer scale. Grounded in years of agent research, our AI is designed with trustworthiness and reliability as core pillars, not afterthoughts. We are supported by tier-1 investors who funded the first generation of AI giants; now they're backing us to build the next: everyday AGI.If you see possibility where others see limits, read on. Build the systems our agents run on. Make them feel like magic, not infrastructure. Our agents are only as trustworthy as the backend behind them. Every API call, every retry, every queued job is the difference between an agent that feels alive and one that feels like a chatbot. You'll own the systems that turn research into product — model orchestration, agent state, partner integrations, the data plane — and you'll make them fast enough, observable enough, and boring enough that the rest of the team can ship without thinking about them. This is for the backend engineer who has been the on-call hero before, who has strong opinions about queues and idempotency, and who wants to design infrastructure for a product where the floor is "millions of devices."Tasks You Will OwnThe agent backend end-to-end — model orchestration, tool-use plumbing, agent state, retries, observability The data plane behind every agent action — Postgres schema, caching, queues, event streams — at OEM scale Production SLAs, on-call, and the reliability bar for everything the agent touches in the cloud Areas Where You Will Assist Research, by shipping their work to real users in days, not quarters — and feeding back what breaks in production Forward-deployed engineers, by giving them backends partners can integrate against without a six-week meetingiOS and Android, by drawing the right line between on-device and cloud so neither side carries the wrong weight Skills You'll Be Expected To TeachHow to design backends for LLM-powered systems where latency, cost, and non-determinism are first-class concerns How to run a production system you'd happily put your name on — observability, incident response, capacity planning Skills You'll Be Expected To LearnThe internals of agentic systems from the people who published the canonical papers on them What it takes to run agent infrastructure at OEM scale, across Samsung, OPPO, Lenovo, and Vertu devices On-device / cloud co-design — when to push compute to the phone and when to keep it on our side Timeline Of Success After 30 days — You've shipped a meaningful change to the agent backend that a user would feel — latency, reliability, or a new capability. You can name the three weakest links in our infra and have started fixing one. You've taken a real on-call shift. After 60 days — You own a major surface of the backend. A research breakthrough has shipped through systems you designed. Other engineers route their hardest backend questions to you. You've set the reliability and observability bar that new services have to meet. After 90 days — Our agent backend is something you'd happily defend to an infra engineer at a top-tier consumer company. You've shaped the architecture for one of our partner launches and have a strong, opinionated plan for what our infra needs to look like to support 10M devices. Compensation Competitive cash and meaningful equity. Top-tier relocation and immigration support. SF, in person.
How To Apply: Send a link to a backend system you've built or operated, with one paragraph on a production decision you'd defend on a whiteboard. Plus your resume, GitHub, or LinkedIn. Every exceptional candidate hears back within 48 hours.

Before you apply

Applying takes about a minute. These four things decide how fast it moves after that.

Your profile is current

It's what we read first. Occupations, seniority and locations matter more than a long history.

Two examples you can talk through

Not a portfolio — just two pieces of work where you can explain the decisions and what you'd change.

A number in mind

What you're on now and what would make you move. We negotiate better when we know both.

Your notice period

Employers plan around it, and it's the question that stalls offers most often.

Once you apply, someone reads it and calls you before anything reaches the employer — usually within two working days.

More like this