OpenAI· Hardware· San Francisco
System Software Engineer, First-Party Hardware
Comp$266K – $445K
Classified Tasks (25)
Automate 0%Augment 68%Human-Only 32%
Augment (17)
AI assists, human decides
Design low-level firmware and system software for first-party AI hardware manageability, including BMC, Redfish, gNMI, firmware update/recovery flows, BIOS/UEFI interactions, platform drivers, and hardware diagnostics.
technical
Develop low-level firmware and system software for first-party AI hardware manageability and health.
technical
Integrate low-level system software with BMC, Linux, firmware interfaces, boot and recovery flows, host and platform drivers, network interfaces, and manufacturing/fleet systems.
technical
Validate low-level system software and produce launch-readiness evidence for first-party AI hardware systems.
operational
Work across BMC, Linux, firmware interfaces, automation infrastructure, boot/recovery, hardware diagnostics, telemetry, host/platform drivers, network software, and manufacturing/fleet readiness to enable platform manageability.
technical
Define software requirements for partner and vendor system software acceptance.
analytical
Review partner and vendor code and artifacts for correctness and readiness.
technical
Reproduce builds from partner and vendor deliverables to verify reproducibility.
technical
Build and run tests, including CI and regression tests, to validate partner and vendor software releases.
technical
Track versions and monitor regressions for partner and vendor software releases.
analytical
Build and maintain automation and continuous integration infrastructure for testing and managing systems in the lab.
technical
Build system health monitoring, telemetry, remote diagnostics, and recovery paths for lab, manufacturing, and production environments.
technical
Develop validation and test automation for board bring-up, rack bring-up, qualification, manufacturing readiness, deployment readiness, and long-term reliability.
technical
Convert engineering releases into manufacturing-ready software recipes, including images, versions, logs, limits, remediation mapping, provisioning hooks, secure artifact handling, and traceable data export.
operational
Produce architecture notes, runbooks, validation records, and decision documents to enable reproducibility, operation, and platform improvement.
administrative
Write and review low-level software and firmware code and associated artifacts.
technical
Build infrastructure and automation to test and manage devices in the lab.
technical
Human-Only (8)
Requires human judgment
Maintain low-level firmware and system software for first-party AI hardware in lab and production environments.
operational
Own the acceptance path for partner-delivered and vendor software releases.
leadership
Push fixes to partner and vendor software and coordinate remediation.
communication
Define and debug hardware management protocols across accelerators, host systems, management controllers, firmware, and platform services using interfaces such as I2C, SMBus, PMBus, PCIe, Ethernet, GPIO, UART, and JTAG.
technical
Debug complex production issues spanning hardware signals, BMC firmware, BIOS/UEFI, kernel drivers, platform services, network topology, PCIe behavior, power, thermals, boot, provisioning, and manufacturing test.
technical
Partner with hardware, firmware, security, networking, infrastructure, manufacturing, operations, and external engineering teams to define software contracts, unblock bring-up, and drive issues to closure.
communication
Guide partner deliverables through acceptance, testing, and launch-readiness workflows.
leadership
Drive platforms from bring-up through production deployment and operational handoff.
leadership
Job description
System Software Engineer, First-Party Hardware | OpenAI Careers ## System Software Engineer, First-Party Hardware Hardware - San Francisco Apply now(opens in a new window) ### **About the Team** OpenAI’s Hardware organization develops silicon and system-level solutions designed for the unique demands of advanced AI workloads. The team is responsible for building the next generation of AI-native silicon while working closely with software and research partners to co-design hardware tightly integrated with AI models. In addition to delivering production-grade silicon for OpenAI’s supercomputing infrastructure, the team also creates custom design tools and methodologies that accelerate innovation and enable hardware optimized specifically for AI. ### About the Role We're seeking a System Software Engineer to join our First-Party Hardware team. In this role, you will design, build, integrate, and validate low-level system software for the manageability and health of OpenAI's first-party AI hardware systems. You will work across BMC, Linux, firmware interfaces, automation infra, boot and recovery, hardware diagnostics, telemetry, host and platform drivers, network software interfaces, and manufacturing and fleet readiness. A major part of this role is owning the acceptance path for partner-delivered system software: defining requirements, reviewing code and artifacts, reproducing builds, building tests, pushing fixes, and producing the evidence needed for launch decisions. This role is hands-on and high-ownership. You will write and review low-level software, debug issues across hardware and software boundaries, build infra and automation to test and manage devices in lab, guide partner deliverables, build validation evidence, and help carry platforms from bring-up through production deployment. *Location: San Francisco, CA (Hybrid: 3 days/week onsite)* *Relocation assistance available.* ### In this role, you will: * Design, develop, and maintain low-level firmware and system software for first-party AI hardware manageability, including BMC software, Redfish services, gNMI telemetry, firmware update and recovery flows, BIOS/UEFI interactions, platform drivers, and hardware diagnostics. * Own integration and acceptance of partner and vendor software releases, including requirements, code and artifact review, reproducible builds, CI, regression monitoring, version tracking, acceptance criteria, and launch-readiness evidence. * Build and maintain automation and CI infra for testing and managing systems in our lab * Define and debug hardware management protocols across accelerators, host systems, management controllers, firmware, and platform services, including interfaces such as I2C, SMBus, PMBus, PCIe, Ethernet, GPIO, UART, and JTAG. * Build system health monitoring, telemetry, remote diagnostics, and recovery paths that make hardware failures diagnosable in the lab, at manufacturing partners, and in production data centers. * Develop validation and test automation for board bring-up, rack bring-up, qualification, manufacturing readiness, deployment readiness, and long-term reliability. * Convert engineering releases into manufacturing-ready software recipes: images, versions, logs, limits, remediation mapping, provisioning hooks, secure artifact handling, and traceable data export. * Debug complex production issues spanning hardware signals, BMC firmware, BIOS/UEFI, kernel drivers, platform services, network topology, PCIe behavior, power, thermals, boot, provisioning, and manufacturing test. * Partner with hardware, firmware, security, networking, infrastructure, manufacturing, operations, and external engineering teams to define software contracts, unblock bring-up, and drive issues to closure. * Produce durable architecture notes, runbooks, validation records, and decision documents that help OpenAI and partner teams reproduce, operate, and improve the platform. ### You might