AI & HPC Infrastructure Engineer

About FirstPrinciples
FirstPrinciples is a research organization building AI infrastructure for discovery in fundamental science. Currently, our work focuses on building systems like Theo, the AI Physicist, which is a domain-specialized system for research in fundamental physics. We’re a fast-growing, remote-first team of builders, researchers, engineers, and thinkers working across Canada, the US, the UK, and expanding globally.

What brings us together is a shared curiosity about how the universe works, and a belief that we can build systems that help us explore it more effectively. We spend our time working on questions that don’t have clear answers, like how to design AI that can reason through scientific problems, and how the scientific process as a whole might evolve. This is work that sits somewhere between creativity and rigorous thinking, and often requires comfort with ambiguity and iteration.

If you’re someone who enjoys tackling big, abstract problems and building the infrastructure that makes ambitious research possible, you’ll likely find the work here interesting.

Why This Role Exists
We’re building the next generation of infrastructure for AI-driven scientific discovery, and we need someone who can help own the systems that make our research and inference workloads reliable, scalable, and fast. This role is about building and operating the compute foundation behind our AI Physicist: Kubernetes clusters, Linux systems, GPU infrastructure, cloud environments, HPC-style compute, deployment workflows, monitoring, and automation. As our workloads grow, we need infrastructure that can support both experimentation and production-like inference across cloud, bare metal, and hybrid environments.

You’ll play a central role in shaping how we run compute at FirstPrinciples. That includes provisioning and managing clusters, improving reliability and observability, reducing operational toil, supporting researchers and engineers, and helping us make practical decisions about when to use managed cloud services, self-managed Kubernetes, Slurm-style systems, or owned hardware. We’re looking for someone hands-on, systems-oriented, and comfortable working in a fast-moving research environment.

You should have strong Kubernetes and Linux fundamentals, good operational instincts, and enough experience with cloud and HPC/GPU infrastructure to help us build toward a robust bare metal and multi-cloud inference platform.

What You’ll Do
Design, deploy, and operate Kubernetes infrastructure for AI inference, research, and engineering workloads

Set up and manage GPU and HPC-style compute environments, including scheduling, utilization, job management, and node-level troubleshooting

Work with systems such as Kubernetes, Slurm or similar schedulers, container runtimes, GPU drivers & libraries (ie; CUDA), storage systems, and observability tools

Build and manage Linux-based compute environments, including provisioning, networking, storage, monitoring, access control, and lifecycle management

Help architect bare metal, cloud, and hybrid infrastructure across AWS, GCP, Azure, or equivalent platforms

Own the reliability and operational health of infrastructure systems, including monitoring, alerting, incident response, capacity planning, and performance tuning

Improve deployment workflows, automation, configuration management, secrets management, and infrastructure-as-code practices

Partner with ML engineers, researchers, and software engineers to understand workload requirements and translate them into practical infrastructure designs

Evaluate tradeoffs between managed cloud services, self-managed Kubernetes, HPC schedulers, bare metal deployments, and multi-cloud architectures

Build tooling, documentation, runbooks, and operational practices that help the team move quickly without making infrastructure fragile or opaque

Balance speed and robustness, knowing when to prototype quickly and when to harden systems for long-term use

Who You Are
Strong infrastructure builder with experience operating production, research, cloud, or high-performance compute systems

Deeply comfortable with Linux administration, including debugging networking, storage, system services, permissions, performance issues, and node-level failures

Experienced with Kubernetes in real environments, including cluster operations, deployments, networking, observability, scaling, and troubleshooting

Comfortable working with cloud infrastructure on AWS, GCP, Azure, or equivalent platforms

Familiar with infrastructure automation and configuration tools such as Terraform, Ansible, Helm, ArgoCD, GitOps workflows, or similar systems

Experienced with GPU-heavy, compute-heavy, or HPC-style workloads, especially in environments involving AI, ML, research computing, or scientific workloads

Able to work across bare metal and cloud environments, and interested in the practical tradeoffs between the two

Comfortable reasoning about resource scheduling, cluster utilization, autoscaling, storage, networking, and observability for distributed workloads

Practical and ownership-oriented; you can take ambiguous infrastructure needs and turn them into working systems

Comfortable collaborating across disciplines, especially with researchers and engineers who may not think in infrastructure terms

Able to operate independently as a senior or strong intermediate contributor, while knowing when to bring others into important technical decisions

Motivated by building foundational systems that make ambitious technical and scientific work possible

Bonus
Hands-on experience with production-grade LLM inference and serving engines, such as vLLM, SGLang, or TensorRT

Experience working at an AI company, ML infrastructure team, research lab, university compute environment, HPC center, or scientific computing organization

Experience supporting model inference, model serving, distributed training, high-throughput batch workloads, or internal ML platforms

Hands-on experience with Slurm or similar HPC schedulers, including job scheduling, resource allocation, queue management, and cluster configuration

Experience operating GPU infrastructure, including NVIDIA drivers, CUDA, container runtimes, scheduling, utilization, and hardware failure modes

Experience with RDMA, InfiniBand, high-performance networking, distributed filesystems (ie. Lustre, BeeGFS), object storage, or storage systems for compute-heavy workloads

Experience with Kubernetes operators, custom controllers, CRDs, or platform tooling for AI/ML workloads

Experience with Prometheus, Grafana, Loki, OpenTelemetry, Datadog, or similar monitoring, logging, and observability tools

Experience with container registries, image optimization, CI/CD systems, deployment pipelines, and secure software delivery

Experience leading engineering operations or infrastructure efforts while remaining hands-on technically

Familiarity with security, access control, secrets management, and reliability practices in production or research environments

What You’ll Get
The opportunity to work on foundational problems at the intersection of AI and physics

A high-trust, low-bureaucracy environment with real ownership

Remote-first work with flexibility in how you structure your day

Exposure to cutting-edge ideas across AI, scientific discovery, infrastructure, and emerging technologies

A culture that values curiosity, depth of thinking, and first-principles reasoning

The chance to shape the compute and inference infrastructure behind advanced AI systems for scientific discovery