Builder mindset: engineers who love building products and systems enjoy solving hard technical problems naturally think about architecture create solutions instead of waiting for instructions. Looking for architects, people with major roles in building complex systems.

Support existing product / speed up development

The company runs one of the largest GPU fleets in the world. The Infra Agent Systems team builds the software systems that power and automate that infrastructure.

They develop production AI agents that diagnose hardware failures, investigate incidents, correlate signals across the fleet, and automate operational workflows. Alongside these agents, we build the platform they run on, including knowledge graphs, retrieval systems, orchestration frameworks, and developer tooling.

You’ll work across two areas:

Infrastructure Agent Systems — Build production AI agents that help operate our GPU fleet by diagnosing failures, investigating incidents, gathering evidence from live systems, and assisting with remediation. These agents are used every day by our infrastructure and datacenter teams through APIs, CLI, dashboards, and Slack.

Core Agent Platform — Build the platform that powers these agents, including knowledge graphs, search and retrieval, orchestration, evaluation, and the tooling that enables agents to reason, act, and continuously improve.

The team is working on something that hasn’t really been done before: building knowledge graphs and self-improving AI agents that understand, operate, and continuously improve large-scale AI infrastructure.

This is an opportunity to work at the intersection of AI agents, distributed systems, infrastructure, and automation, solving challenging engineering problems with real production impact. There’s an enormous amount to build, learn, and shape as we define the future of autonomous infrastructure.

Why this Role

You’ll work on two hard problems at the same time: making AI agents trustworthy enough to operate production infrastructure, and building the knowledge, retrieval, and distributed systems that make those agents effective.

You’ll have the opportunity to build foundational systems from the ground up, work on infrastructure at massive scale, and help define how self-improving AI agents operate real-world AI infrastructure.

Responsibilities:

• Design and build production AI agent systems that diagnose, investigate, and remediate infrastructure issues across one of the world’s largest GPU fleets.
• Build the distributed services, orchestration framework, knowledge graph, and retrieval systems that power infrastructure agents.
• Develop fleet intelligence systems that combine telemetry, infrastructure state, operational knowledge, and historical incidents to help agents make better decisions.
• Integrate with observability, incident management, ticketing, fleet inventory, source control, chat, and internal infrastructure systems through well-designed APIs.
• Own services end to end, including architecture, implementation, testing, deployment, observability, and production operations.
• Improve agent performance through evaluations, retrieval improvements, better tools, and production feedback loops.
• Turn what agents learn in production into reliable, reviewed software and automation.

Requirements

• 5+ years of experience building production backend systems, distributed systems, or infrastructure platforms.
• Strong expertise building AI agentic systems, with a deep understanding of their internal design and how to make them effective.
• Experience with Kubernetes, GitOps such as ArgoCD, infrastructure-as-code, and cloud platforms, especially in the context of AI systems.
• Strong systems design skills and experience owning significant systems from design through production.
• Strong backend engineering experience, including API design, service boundaries, data modeling, and integrations across complex systems.
• Comfortable working across languages such as Go, TypeScript, Python, or Rust.
• Significant personal ownership of core agentic fundamentals.
• Experience building knowledge graphs and working through the associated problems.
• Evidence of significantly improving agent accuracy and effectiveness.
• Ability to go beyond RAG-based implementations and AI-enabled product development.

Nice to have:

• GPU infrastructure, datacenters, bare-metal systems, hardware failure modes, BMC/IPMI, or cluster schedulers
• Graph databases
• Event-driven systems and messaging platforms such as NATS or Kafka
• Observability platforms such as Prometheus and Grafana
• Building evaluation frameworks or improving the quality and reliability of LLM-powered systems

AI Native Cloud purpose-built for AI engineers. AI application developers get high-performance inference that scales reliably, fine-tuning and reinforcement learning for creating frontier-level specialized models, and pre-training at massive scale for fully custom intelligence, all around a marketplace of leading open models that teams can run, adapt, and own. Trusted by Cursor, Decagon, ElevenLabs, Salesforce, and Zoom, serves 400+ trillion tokens a month.

to be discussed

• Slack — main channel for daily communication, coordination, onboarding questions and interview discussions.
• Direct collaboration with principal engineers and team leads.

10+