Infrastructure Software Engineer, Normal Computing

• This is a highly cross-functional role for someone who enjoys ambiguity, cares about clean abstractions, and wants to help shape how a frontier AI company builds and operates production systems. Strong engineering judgment and ownership matter more than rigid specialization
• 4+ years of experience in infrastructure software, backend infrastructure, production infrastructure, platform engineering, distributed systems, or related areas
• Strong software engineering fundamentals, including backend programming, APIs, data modeling, concurrency, debugging, and testing
• Experience building or operating production services where reliability, observability, and maintainability matter
• Practical experience with Docker and Kubernetes, including debugging containerized workloads, deployments, networking, resource limits, and lifecycle issues
• Comfort working with persistence systems such as Postgres, Redis/Valkey, object storage, or similar production data stores
• Experience building orchestration systems, job schedulers, workflow engines, sandboxes, developer platforms, or distributed execution systems
• Experience designing internal APIs and developer-facing abstractions that other engineers can use confidently
• Strong systems thinking: you can reason about state machines, failure modes, retries, queues, leases, scheduling, and long-running workflows
• Pragmatism in fast-moving environments: you know when to improve an abstraction, when to delete one, and when to ship the simple version
• Ownership mindset: you care whether the systems you build work in production and are usable by other engineers
• Clear communication and good technical judgment across product, AI, and infrastructure boundaries

Desirable

• We’re looking for an Infrastructure Software Engineer to build the production systems behind Normal’s AI products
• This is an application engineering role focused on infrastructure-shaped software: orchestration services, execution runtimes, internal APIs, persistence layers, observability, and developer experience. You’ll help define the runtime layer for a new class of AI products: systems where agents execute long-running work, coordinate across distributed environments, interact with code and tools, and need to be reliable enough for real customer workflows
• This role sits between product engineering, AI engineering, and platform engineering. You will not primarily be managing Terraform, Helm charts, CI/CD, or company-wide SaaS infrastructure. Instead, you’ll own the application-level infrastructure that powers long-running AI workflows: session lifecycle, sandboxed execution, workload orchestration, persistence, observability, reliability, and the internal interfaces other engineers build on
• The systems you build will be used directly by product, AI, research, and platform teams as new capabilities move from early ideas into production. Developer experience matters: APIs should be understandable, failure modes should be debuggable, and abstractions should make the right thing easy
• On any given day, you might design the runtime architecture for a new AI product capability, build the orchestration layer for long-running autonomous workflows, improve how workloads are scheduled and isolated across distributed environments, or create the systems abstractions that let engineers turn ambitious AI prototypes into reliable production products
• Build and maintain production software infrastructure for Normal’s AI products, especially orchestration, execution, and runtime systems
• Design internal backend services and APIs used by product engineers, AI engineers, execution services, and other internal systems
• Improve the operational maturity of rapidly evolving systems through better state management, failure handling, metrics, tracing, and debugging tools
• Work with Kubernetes-backed execution environments, including container lifecycle, scheduling behavior, autoscaling, resource isolation, and runtime reliability
• Build developer-facing tools and abstractions that make it easier for other engineers to use and extend the systems you own
• Turn promising prototypes into durable production systems by designing clear abstractions, hardening critical paths, and creating operational patterns that scale with the product
• Collaborate closely with product, AI, research, and platform engineers to define the right interfaces between product features, AI workloads, and production infrastructure
• Lead design discussions for core runtime and orchestration systems, including API boundaries, state management, execution models, and operational tradeoffs