🏗️ AI Infrastructure Engineer

Confidential AI builds the confidential computing stack for AI. We sell private inference and confidential GPU VMs.

Every prompt, CPU, and GPU is secured in a hardware-encrypted Trusted Execution Environment (TEE) with NVIDIA confidential computing, orchestrated and scaled in our confidential, verifiable fabric. You will own the inference infrastructure that makes that work, from silicon to Kubernetes, on our hardware and others'.

Our team is small and high-agency. You will own principal pieces of the puzzle.

What You'll Work On

You'll design, scale, and optimize AI inference infrastructure at data-center scale. You'll work across the entire inference stack — from GPU kernels, memory, KV cache, and inference engines to networking, orchestration, and fleet-wide scheduling. You know what breaks at 1,000 GPUs that did not break at 8. You know how to make every GPU, node, and cluster deliver more tokens.

This role is production inference, end to end. Design it. Deploy it. Scale it. Monitor it. Optimize it.

• Design, deploy, and operate GPU clusters (H100, H200, B200, B300) as Confidential VMs on AMD SEV-SNP and Intel TDX hosts.
• Scale Confidential Kubernetes (C8s) across bare metal, Azure, GCP, AWS, and neoclouds.
• Build the control plane for scheduling, autoscaling, and orchestrating inference inside TEEs.
• Bring up multi-GPU CVMs with encrypted NVLink and protected PCIe. Work with NVIDIA, AMD, and Intel on next generation confidential computing hardware
• Run inference fleets of SGLang, vLLM, and NVIDIA Dynamo.
• Deploy dedicated inference deployments for large enterprise customers.
• Own capacity planning, fleet health, incident response, and the runbooks that go with them.
• Automate everything. Terraform, Helm, CVM image pipelines, attestation-gated secret distribution.

Requirements

Confidential AI values these traits above all others:

• Consistent: You ship daily.
• Relentless: You find a way or make a way.
• Curious: You thrive in the unfamiliar. New problems energize you.
• Collaborative: People want to work with you. You communicate clearly.

Qualifications

None. Really. We don't care where, or if, you went to school.

Bonus Experience

It's a bonus, not a requirement, if you have experience with any of the following:

• Operating data center scale GPU clusters
• Scaling inference endpoints at an AI lab or inference provider
• Multi-node inference serving: tensor and pipeline parallelism, NCCL tuning
• Kubernetes internals: schedulers, device plugins, CNI, Kata Containers
• Bare metal provisioning and firmware: UEFI, PXE, BMC, IPMI
• InfiniBand, RoCE, NVLink, GPUDirect
• Confidential computing: AMD SEV-SNP, Intel TDX, NVIDIA CC
• Golang, Rust, and/or Python
• Terraform, Helm, Nix