Software Engineer - Infrastructure

fal is the generative media ecosystem powering the next generation of AI products. We build the infrastructure, tools, and model access that teams need to move from idea to production, and do it at scale without compromise. For developers and enterprises, fal is the foundation that makes generative media not just possible, but practical: a unified platform where high-performance inference, orchestration, and observability come together to unlock new categories of AI-native products. As generative media reshapes industries across a market projected to grow by hundreds of billions over the next decade, fal is becoming the ecosystem that ambitious teams build on. You are a hands-on engineer who builds the software and processes that keep a large fleet of GPU servers healthy and productive. You write systems and tooling for managing 1000s of servers including provisioning, health monitoring, error detection, and recovery — and when something breaks that automation can’t fix, you drive resolution with partners. Key Responsibilities Build and maintain Python fleet tracking system that manages the full lifecycle of servers including contracting and procurement, target use, pricing, availability, health, RMAs, etcBuild server management tooling that automates provisioning, health checks, GPU diagnostics, recovery and alertingCreate and maintain metrics, dashboards, and alerting for hardware health across the fleet (GPU errors, disk failures, network issues, thermals)Leverage AI to an extreme level to build tools and automate alerting and recoveryImplement and enforce OS-level security: hardening baselines, SELinux/AppArmor policies, SSH key management, vulnerability scanning, and compliance automationManage and optimize distributed and local storage systems supporting model weights, checkpoints, and ephemeral scratch: NVMe arrays, NFS, parallel file systems, and object storageTune Linux systems for AI workloads: kernel parameters, NUMA topology, CPU pinning, hugepages, I/O schedulers, and GPU driver stack optimization (NVIDIA drivers, CUDA, container runtimes)Develop a suite of automated error detection and recovery processesWork with partners to solve technical issues Requirements 3+ years experience managing bare-metal and cloud based server fleets at scale (100+ nodes)Strong software engineering skills in Python; you write production tooling, not scriptsDeep Linux systems knowledge: boot process, kernel tuning, networking, storage, systemd, cgroups, namespaces, performance profilingStrong experience with configuration management and infrastructure-as-code: Ansible, Terraform, cloud-initSolid understanding of storage technologies: LVM, RAID, NVMe, NFS, Lustre or GPFS, and Linux I/O stack tuningFamiliarity with hardware diagnostics and failure modes (GPUs, NVMe, NICs, memory)Experience building internal tools or dashboards for infrastructure visibilityExcellent communication and ability to drive technical decisions across teamsSelf-starter who executes quickly, takes ownership, and constantly seeks improvement Nice to have Familiarity with network configuration and diagnostics (VLAN, VXLAN, ECMP, BGP, tcpdump)Experience with NVIDIA GPU infrastructure: driver management, health monitoring, DCGM, NVLink/NVSwitch diagnostics, RDMA, InfiniBand/RoCEv2Experience with AMD GPUsExperience with bare metal and VM provisioning (PXE/iPXE, Kickstart, libvirt, Qemu/KVM)Experience with compliance frameworks relevant to cloud providers (SOC 2, ISO 27001) Compensation $180,000-250,000 plus equity + benefits Location San Francisco, CA (we are open to remote in the US for Senior and Staff levels) What we offer At Fal Interesting and challenging workA lot of learning and growth opportunitiesWe are offering relocation assistance to San Francisco. We offer relocation assistance to San Francisco. Health, dental, and vision insurance (US)Regular team events and offsites