Platform Engineer

Role 1 - Core Platform Engineer (L1, breadth-first) The engineering first line of defense: incident response, triage, reliability, and automation across the full infrastructure stack. Day-to-day: Runs incident response drills, post-mortems, and root cause analysis; learns from past incidents to prevent recurrence. Starts the day reviewing overnight alerts and system performance metrics, triaging anomalies. Participates in team stand-ups on projects, incidents, and daily priorities. Automates routine processes, analyzes system logs, and builds tools to strengthen monitoring. Works alongside software engineers advising on resilient-code best practices and reviewing changes pre-deployment. Maintains high SLIs/SLOs; documents work and shares insights with a customer-centric mindset. Must have: Architecture, design patterns, reliability, and scaling of new and existing systems. Incident command experience - driving RCA, coordinating cross-functional teams, ensuring corrective-action follow-through. Observability built from the ground up - defining SLOs/SLIs, closing monitoring gaps, alerting strategies that catch failures before customers do. Linux kernel internals - scheduler, memory allocation, driver subsystems. High-quality code in at least one language (Python, Go, or similar). System-level debugging - kdump, kernel panic analysis. IaC (Ansible, Terraform, Kubernetes) and CI/CD (GitLab CI, AWX, etc.) for bare-metal or cloud infrastructure. TCP/IP and network programming. Distributed storage systems - object, block, and/or file storage paradigms. Strong communication skills. Nice to have: Hardware and GPU troubleshooting. OVN/OVS-based networking stack exposure. Sourcing note: This is a deep SRE profile, not a pure generalist. The kernel-internals and system-level debugging bar is real and higher than a typical "L1" label implies - screen for genuine engineering depth, not helpdesk/NOC-tier breadth. Role 2 - Platform Engineer (L2, depth-first) Specialized domain expert embedded in a single foundation team: Storage, Compute, or SDN. Owns reliability, performance, scalability, and operational excellence for that one pillar. No cross-vertical work. Must have: Deep expertise in one domain (Storage / Compute / SDN). SRE as a fundamental minimum - has set up observability, done alert management, and worked with SLIs/SLOs within that domain. Ability to code the infrastructure: extend out-of-the-box product parameters to build observability layers, and tie SLAs/SLOs to business KPIs. Domain specifics: o Storage: block, blob, and file storage; distributed storage; performance diagnostics and data-path optimization. Critical screening filter - operator vs. SRE: Crusoe explicitly does not want another "operator" (storage admin doing patching, installs, upgrades). Screen hard for SRE substance (observability built, alerting, SLI/SLO ownership), not just domain tenure. Role 3 - Platform Engineer (L2, depth-first) Specialized domain expert embedded in a single foundation team: Storage, Compute, or SDN. Owns reliability, performance, scalability, and operational excellence for that one pillar. No cross-vertical work. Must have: Deep expertise in one domain (Storage / Compute / SDN). SRE as a fundamental minimum - has set up observability, done alert management, and worked with SLIs/SLOs within that domain. Ability to code the infrastructure: extend out-of-the-box product parameters to build observability layers, and tie SLAs/SLOs to business KPIs. Domain specifics: o Compute: Linux systems, KVM/QEMU, Cloud Hypervisor, kernel tuning, CPU/memory/VM optimization. Critical screening filter - operator vs. SRE: Crusoe explicitly does not want another "operator" (storage admin doing patching, installs, upgrades). Screen hard for SRE substance (observability built, alerting, SLI/SLO ownership), not just domain tenure. Role 4 - Platform Engineer (L2, depth-first) Specialized domain expert embedded in a single foundation team: Storage, Compute, or SDN. Owns reliability, performance, scalability, and operational excellence for that one pillar. No cross-vertical work. Must have: Deep expertise in one domain (Storage / Compute / SDN). SRE as a fundamental minimum - has set up observability, done alert management, and worked with SLIs/SLOs within that domain. Ability to code the infrastructure: extend out-of-the-box product parameters to build observability layers, and tie SLAs/SLOs to business KPIs. Domain specifics: SDN: OVS/OVN, network virtualization, high-performance networking, NIC tuning. Critical screening filter - operator vs. SRE: Crusoe explicitly does not want another "operator" (storage admin doing patching, installs, upgrades). Screen hard for SRE substance (observability built, alerting, SLI/SLO ownership), not just domain tenure.