Director Site Reliability Engineer, AI Infrastructure

The role

This person will build and lead the Site Reliability Engineering function from the ground up. They will own infrastructure spanning colocation, on-premises lab clusters, cloud environments, and platform services.

What you will do

• Build and lead the SRE function from scratch and hire a team of 1–3 engineers.
• Lead the Data Center & Lab Technician team.
• Own 24×7 reliability across colocation, on-premises lab clusters, cloud, and customer-facing platform services.
• Establish SRE processes including SLIs, SLOs, error budgets, on-call rotations, and incident/RCA frameworks.
• Own the observability stack.
• Drive IaC-first automation and build self-healing infrastructure.
• Own FinOps and capacity planning across cloud, colocation, and on-premises environments.
• Lead the migration from ad-hoc JBOD storage to an enterprise-grade shared storage platform.

Required qualifications

• Bachelor’s or Master’s degree in Computer Science, Electrical Engineering, or a related field.
• 15+ years in SRE, infrastructure engineering, or production engineering.
• 5+ years leading SRE or infrastructure engineering teams, including experience building or significantly rebuilding a function.
• Track record of establishing SRE as a discipline in an organization that lacked it.
• Deep Linux systems expertise, including bare-metal operations and enterprise shared storage.
• Experience operating colocation and on-premises hardware at scale.
• Hands-on production experience with Infrastructure as Code (Terraform, Ansible).
• Experience with Kubernetes cluster operations.
• Experience owning an observability stack (Prometheus, Grafana, Datadog, Splunk, or equivalent).
• Strong Python and/or Go skills.
• Ability to communicate infrastructure health and operational risk clearly to senior leadership and non-technical stakeholders.
• Comfort operating in a high-ambiguity, low-process environment.

Nice to have

• Experience operating customer-facing infrastructure or platform services.
• Knowledge of high-speed interconnect fabrics such as InfiniBand, RoCE, or NVLink.
• Experience with HPC job schedulers (Slurm, LSF, or equivalent).
• Multi-cloud hybrid operations across AWS, Azure, and GCP alongside on-prem and colo.
• FinOps expertise.
• Familiarity with ITIL or an equivalent structured incident, problem, and change management framework.
• Published technical writing, conference talks, or open-source contributions in reliability, observability, or HPC infrastructure.

About the company

d-Matrix designs and manufactures purpose-built AI inference silicon. They have pioneered a memory-centric architecture that brings compute directly into memory.