Infrastructure Engineer Intern

Dedalus Labs is an AI research neolab building infrastructure for AI agents, including distributed storage, virtualization, orchestration, networking, scheduling, and runtime infrastructure. The Infrastructure Engineer Intern will work in person with engineers on production infrastructure, taking ownership of meaningful projects involving distributed systems, reliability, performance, observability, and fault tolerance.

Responsibilities:

• Distributed infrastructure for large-scale AI agent workloads
• Persistent compute and distributed storage systems
• Scheduling and orchestration platforms
• Virtualization, containerization, and sandboxing infrastructure
• Reliable multi-tenant cloud systems
• Internal developer platforms and infrastructure tooling
• Performance, reliability, observability, and failure recovery
• Production systems operating under real-world latency and fault-tolerance constraints
• Building a component of a distributed storage system for persistent agent state
• Improving the reliability or recovery behavior of a production service
• Designing scheduling infrastructure for concurrent, long-running agent workloads
• Investigating a bottleneck across storage, networking, scheduling, or runtime layers
• Building internal tooling that makes infrastructure easier to operate and debug
• Designing tests that simulate partial failures, machine loss, or degraded networks
• Improving observability for distributed workloads
• Prototyping a new isolation, caching, or orchestration mechanism
• Reading a systems paper, implementing part of it, and comparing its tradeoffs with our architecture