HU01 - AI Infrastructure Engineer

About the Project

We are building the enterprise cloud platform for running AI agents securely at scale. A useful way to think about this role:

Your job is to build the infrastructure to host OpenClaw for 10,000 non-technical users in an extremely secure environment. If you’ve used, deployed, or are obsessed with OpenClaw, you’ll immediately understand the challenges:

• Agents executing code
• Accessing tools and data
• Running in parallel across users
• Handling untrusted workloads safely This is not traditional backend infrastructure. This is about securely running autonomous systems that can execute arbitrary code. The platform must support:
• Multi-user environments
• Strong isolation between users and organizations
• Secure execution of agent-generated code
• Scalable orchestration across cloud providers (AWS, GCP, Azure)

---

Role and Responsibilities

Secure Agent Execution (Core Problem)

• Design infrastructure for running untrusted agent code safely
• Build strong isolation between users and workloads
• Explore and implement microVM-based approaches (e.g., Firecracker-style isolation)
• Understand tradeoffs vs containers (e.g., Kubernetes)

This is the hardest and most important part of the role.

---

Infrastructure & Platform

• Design multi-cloud architectures (AWS, GCP, Azure)
• Build Infrastructure-as-Code using:
  • Terraform (preferred) or Pulumi
• Create simple, fast deployment systems:
  • Dev / staging / prod
  • Multi-tenant isolation
  • Identity and access control

---

Execution Systems

• Build systems for:
  • Ephemeral workloads
  • Async and event-driven execution
  • Long-running agent processes
• Optimize for:
  • Security
  • Speed
  • Cost

---

Platform Services

• Authentication (SSO, RBAC)
• Secrets and key management
• Networking (VPCs, private connectivity)
• Secure service communication

---

Observability & Reliability

• Logging, metrics, tracing
• Alerting and incident response
• Cost monitoring

---

Internal Developer Platform

• CLI tools / SDKs
• Deployment abstractions
• Environment bootstrapping

---

How We Work (Read Carefully)

This is a high-intensity startup environment.

• We optimize for speed over perfection
• MVPs should be extremely simple
• Overengineering is a negative signal
• We expect you to ship fast and iterate constantly We also expect:
• Heavy use of AI coding tools
• Comfort with AI generating 90%+ of your code
• You are building continuously, not occasionally This is not the right role if:
• You prioritize work-life balance right now
• You prefer long design cycles over shipping
• You default to large, complex architectures early

---

What We’re Looking For

Must-Haves

• 4+ years in cloud / platform / DevOps engineering
• Strong experience with:
  • Terraform (preferred) or Pulumi
• Deep experience in AWS, GCP, or Azure
• Strong coding skills (Python, Go, or TypeScript)
• Experience building real production systems

---

Critical Experience

• Designing systems for:
  • Untrusted code execution
  • Sandboxing / isolation
• Strong understanding of:
  • Networking (VPCs, isolation boundaries)
  • IAM and security models
  • Distributed systems

---

Mindset (This Is What We Actually Filter For)

• You have:
  • Deployed OpenClaw yourself, or
  • Strong interest in doing so immediately
• You think in:
  • “What is the simplest thing that works securely?”
• You:
  • Ship fast
  • Avoid overengineering
  • Iterate aggressively

---

Execution Signals We Value

• You’ve built and shipped projects on your own
• You use AI to write most of your code
• You are comfortable moving from idea → deployed system in days, not months

---

Nice-to-Haves

• Experience with microVMs / Firecracker
• Experience with secure sandboxing systems
• Multi-tenant SaaS architecture experience
• Kubernetes (with awareness of its limitations for isolation)
• Familiarity with agentic systems / LLM platforms

---

Ideal Candidate

• Has tried to run something like OpenClaw at scale and felt the pain
• Understands why containers alone are not enough for isolation
• Builds fast, breaks things, and replaces them quickly
• Is obsessed with AI-assisted development
• Wants to work intensely and move fast

---

Why Join Us?

• Work on one of the hardest problems in AI infra:
  • secure execution of autonomous agents
• Define how AI systems run in real enterprise environments
• High ownership, zero bureaucracy
• Build fast, ship fast, learn fast