Platform Engineer

📋 Job Overview

We're looking for a Platform Engineer to own the infrastructure that puts long-horizon AI Agents inside America's factories. This includes the deployment architecture, observability stack, and industrial connectivity layer that let our agents run anywhere from our cloud to a locked-down plant network. This role is core to how Nexus ships. Every agent investigation, every customer go-live, every on-prem deployment runs on the systems you'll build. Manufacturing doesn't forgive flaky software: our customers run around the clock and their factories collectively ship billions of dollars of product a year.

🏢 About Nexus

Nexus is building the Self-Improving OS for manufacturing. Our platform seamlessly embeds Industrial Agents into decades-old equipment and proprietary industrial control systems to monitor production lines, trace every loss to its root cause, and autonomously act. We're growing quickly, with customers ranging from Fortune 100 manufacturers to startups building cutting-edge supply chains in the United States. Our team is comprised of manufacturing and software experts from Tesla, Distyl, and Convoy, and we're backed by A*, a16z, BoxGroup, and more.

🎯 The Role

You'll work across a few tightly coupled areas. We're flexible on entry points and will match scope to your strengths.

✅ Key Responsibilities

• Define how Nexus ships into wildly different environments: our cloud, customer VPCs, and on-prem servers sitting inside plant networks. You'll own the architecture and processes that make \"deploy to a factory\" as routine as \"deploy to prod.\" Most of our deployments go live in under two weeks - your work is what keeps that true as we scale.
• Architect and run the stack that tells us what every agent is doing, across all three layers: application (agent traces, evals, logs), platform (workflow runner, backend services), and infrastructure (AWS, customer VPCs, on-prem). When something misbehaves inside a customer's plant, your telemetry is how we know before the customer does.
• Build the integration substrate that connects agents natively to every layer of the factory stack - from proprietary control-system file formats and SDKs on up.
• Build and refine the containerization and internal networking stack that lets one platform run identically in AWS and inside a plant DMZ - secure, updatable, and observable even where internet access isn't a given.

📌 Required Qualifications

• 2+ years building and operating production systems, with real ownership of uptime. You know how scalable systems should be architected, built, and maintained.
• Exceptional software development skills, from greenfield to large existing codebases. You can implement systems in about one-third the time most competent people think possible.
• Strong fundamentals in containers, networking, and cloud infrastructure. You can reason about how data and traffic flow through a distributed system, and what breaks first.
• You use AI aggressively in your own engineering workflow. You've tried a bunch of AI developer tools and have strong opinions about what actually works.
• You communicate clearly with teammates, partners, and customers, and you'll respectfully challenge (and be challenged) so that only the best ideas get executed.

⭐ Desirable Experience

• You've shipped agentic or LLM systems to production and think clearly about evals, reliable agent loops, tool design, and context management.
• You've worked with industrial or OT systems (PLCs, SCADA, historians, MES, UNS, etc.) or in other secure, regulated, or air-gapped environments.
• You've built internal tooling consumed by both humans and agents.

🎁 Benefits

Competitive

🛂 Visa & Eligibility

This role is based in San Francisco. Occasional travel may be required for events and customer visits.