About Prem

Prem AI is building the infrastructure layer for private, sovereign AI. Our platform lets enterprises run AI workloads where data is encrypted end-to-end, processed inside hardware-sealed Trusted Execution Environments, and where privacy is enforced by silicon — not a policy document. We're backed by top-tier investors and are emerging from stealth in summer 2026.

About the Role

We are a fast-paced, high-performance team dedicated to shipping world-class products, focusing on disciplined execution and accountability. We are looking for a Platform Engineer to join our infrastructure team, working directly alongside our Head of Infrastructure to own and evolve the confidential compute platform that powers our API.

The developer-facing product is an OpenAI-compatible API where data is encrypted end-to-end and processed exclusively inside hardware-sealed Trusted Execution Environments (TEEs), also called Confidential Virtual Machines (CVMs). Privacy is enforced by hardware — not promised in a policy document. Models are self-hosted; client data never exists in plaintext outside the user's device and the sealed enclave.

The platform runs on on-premise hardware and public clouds. The stack spans cloud infrastructure, low-level systems work, cryptography, and the API/SDK layer that developers actually touch. You will not be siloed — you will be the second pair of hands and eyes on production.

What You'll Do

• Kubernetes Ownership: Be the Kubernetes maintainer. Own cluster operations end-to-end: upgrades, capacity, networking, RBAC, manifests/Helm charts, ongoing operations and maintenance, with operational and governance guidance from the lead. You make the calls; the lead is the sounding board, not the bottleneck.
• Cloud Infrastructure: Run and evolve infrastructure on AWS, GCP, Azure, and others, leveraging each cloud's confidential compute primitives (Nitro Enclaves, Confidential VMs, Confidential Space) to scale the API and to host adjacent, non-API workloads.
• On-Premise Operations: Operate and extend our on-premise confidential compute footprint: bare-metal provisioning, network topology, hardware lifecycle.
• Observability: Own observability for the on-premise side of the platform. Design and operate the metrics, logs, and traces pipeline for bare-metal and CVM workloads where standard cloud-native observability falls short. Make the on-prem fleet as legible as a managed cluster.
• Infrastructure as Code: Maintain provisioning and configuration via Ansible and IaC (Terraform or equivalent): reproducible, code-reviewed, no snowflake servers.
• Platform Hardening: Harden the platform: OS hardening, network segmentation, secret rotation, and reproducible images.
• Backend & API Development: Write and maintain backend services and the API gateway in Node.js / TypeScript. Contribute to performance-critical enclave-side code (we use Rust, willingness to learn is fine if you bring strong systems fundamentals).
• Attestation & Cryptography: Contribute to the attestation pipeline and the cryptographic plumbing that connects clients, enclaves, and our API gateway.
• Runbook Discipline: Enforce the runbook discipline: Every recurring operation, every incident class, every non-obvious recovery procedure gets a runbook. You write them, you keep them current, you push back when work ships without one. If something breaks at 3am, the next person on call does not guess.
• Developer Experience: Help define SLOs and engage with developer users: triage SDK issues, reproduce bugs, ship fixes.

What We’re Looking For

• Linux: Strong Linux sysadmin background. You are comfortable on the shell, you understand systemd, networking (iptables/nftables, TLS, routing), file systems, and how to diagnose a misbehaving Linux box without panicking.
• Kubernetes: Production Kubernetes experience as a maintainer, not a user. You have upgraded clusters, debugged CNI/DNS issues, written manifests/Helm charts, set up RBAC and network policies, and recovered a cluster that didn't want to be recovered.