Founding Engineer - Platform

uRun; San Francisco (CA); On-site; USD 140,000 - 180,000; Full time

Benefits offered by this job

Competitive salary and equity

Full health, dental, and vision coverage

401(k) retirement savings

Flexible spending accounts

Paid time off

Access to top-tier AI tools

Hardware such as MacBook Pro and AirPods

Job summary

uRun, located in San Francisco, is looking for a skilled engineer to design and operate a scalable low-latency infrastructure for real-time AI inference. You will build platforms that support interactive workloads, handling GPU-constrained and bursty scenarios while ensuring reliability and performance at scale.

With over 7 years of engineering experience, you should have deep Kubernetes expertise and a strong background in cloud services. This role offers competitive salary, health benefits, and equity in an early-stage AI company.

Qualifications

• 7+ years as an engineer with experience in large-scale production systems.
• Deep understanding of Kubernetes, particularly with GPU-heavy clusters.
• Solid knowledge of cloud services like AWS, GCP, or Azure.

Responsibilities

• Design, operate, and evolve the cloud-native platform for real-time inference.
• Ensure reliability and performance at scale with observability standards.
• Collaborate with ML teams to optimize workflows for low-latency tasks.

Skills

Cloud-native platform design

Kubernetes expertise

Observability practices

Latency optimization

Python/Go/TypeScript programming

Mentoring

GPU-optimized workloads

Real-time systems

Tools

Prometheus

Grafana

AWS

GCP

Azure

Terraform

Pulumi

Job description

The problem we saw

AI inference today is slow, expensive, and stateless. Send a query, wait, get a response, reset. That's fine for batch — but AI is becoming interactive, and interactive means inference has to respond instantly, hold context across a session, and be steerable in real time.

Nobody had built an infrastructure that does all three at once. The bottleneck isn't the models. It's the runtime underneath them.

What we're building to fix it

uRun — Universal Runtime is the layer that makes real‑time, stateful inference possible. Our platform lets AI respond instantly, hold context across a session, and be directed as it runs.

We prove it through the hardest problem in the stack: real‑time AI video generation. Not pre‑rendered clips. Not queued jobs. Live, steerable, continuous video that responds as you speak. Solve that, and the rest of the inference stack follows, and that's what we've done. We're an infrastructure company; we build the layer model labs, builders, and research teams ship on top of.

Where you come in

You’ll design and own the scalable, low‑latency infrastructure that powers uRun's real‑time inference runtime, the platform that makes live, interactive, multi‑user AI workloads possible.

This is not classic ops or cloud management. You’ll be deep in the AI runtime itself, not just keeping VMs up. Latency, frame rate, and interactive quality of service are first‑class platform properties, and they're yours to own. The workloads are GPU‑constrained, memory‑bound, and bursty, not stateless web backends, so you’ll often write platform features, custom controllers, and scaling logic rather than only operating commercial tooling.

You’ll report directly to our founder, Keegan McCallum, and set the technical direction the engineering organisation grows around.

What you’ll actually be doing day‑to‑day

• Design, operate, and evolve the cloud‑native platform that runs uRun's real‑time inference and video runtime, Kubernetes, GPU‑heavy workloads, and streaming pipelines


• Own observability, reliability, and performance at scale: SLO‑driven capacity, autoscaling, failover, and cost‑efficient GPU provisioning


• Build and maintain the platform primitives that product and ML teams depend on, service meshes, deployment pipelines, secrets and credential management, and configuration‑as‑code


• Partner closely with ML and video‑workload engineers to optimise for low‑latency inference, memory‑bound workloads, and streaming data flows


• Define and champion platform standards for security, observability, and incident response, drawing on SRE‑style practices


• Mentor and unblock other engineers, and act as a technical leader on architecture, trade‑offs, and long‑term platform evolution

What skills you need for the journey

• 7+ years as an engineer, with a proven track record architecting and owning large‑scale production systems


• Deep Kubernetes expertise, including GPU‑heavy clusters (NVIDIA tooling, autoscaling on GPU nodes) and service‑mesh patterns


• Strong cloud and infrastructure‑as‑code: AWS, GCP, or Azure; Terraform, Pulumi, or equivalent; networking and security (VPC, IAM, API‑gateway‑style routing)


• SRE‑style thinking and observability depth: Prometheus/Grafana, OpenTelemetry, distributed tracing, SLOs, incident response, and post‑mortems


• Proficiency in at least one of Python, Go, or TypeScript/Node.js for platform tooling, automation, and glue code


• Experience with streaming or real‑time systems: WebRTC, low‑latency video pipelines, or comparable latency‑sensitive workloads. This is central to the role, not a bonus


• A track record of mentoring engineers and influencing cross‑functional teams

Things that will give you an edge

• Hands‑on experience with GPU‑constrained, memory‑bound, or bursty workloads


• Experience writing custom Kubernetes controllers, scaling logic, or other platform features in‑house


• Early‑stage startup experience: owning ambiguous problems end‑to‑end and setting technical direction with limited scaffolding

What you’ll get in return

• Competitive salary and meaningful equity in an early‑stage AI infrastructure company. The band above is our target; for an exceptional candidate we’ll go higher. Equity is real, you’re early, and the grant reflects that.


• Health, dental, and vision — full coverage


• 401(k) — company‑supported retirement savings


• FSA/HSA — flexible spending accounts for healthcare costs


• Paid time off — we trust you to manage your time


• Top‑tier tooling — access to the best AI tools available: Claude, Codex, Kimi, and whatever else helps you move faster


• MacBook Pro and AirPods — the hardware you need, on us

How we work (and what that feels like day‑to‑day)

We build the stage, not the show. We're an infrastructure company, a developer‑tools company, and a production partner for model labs, and focus is a deliberate choice we've made and hold to.

Day‑to‑day, that means a small team, a high bar, and real ownership. You won’t wait for permission or inherit a backlog of someone else’s decisions, in a founding security role, the function is what you make it.

It also means ambiguity: priorities shift, not everything is documented, and you’ll often be the person who decides what “secure enough, for now” means. That suits some people and not others, and we’d rather you know that before you apply.