Hardware Lab Manager

Anthropic's mission is to create reliable, interpretable, and steerable AI systems. The Foragers team is a new Field Services team within Anthropic's physical infrastructure organization — the people who get hardware into racks, gear commissioned, and sites handed over. This role is the lab-focused counterpart to traveling field engineers: instead of covering sites across the country, you own Anthropic's Bay Area hardware labs from end to end as the first dedicated lab manager, shaping the role to scale from the current lab to much larger ones.

• Run the physical lab across both sites: receiving and staging, rack-and-stack, cabling, power distribution and load planning, cooling, floor layout
• Bring up new server and accelerator platforms from crate to a machine someone can SSH into: firmware and BIOS/UEFI setup, out-of-band management, OS install, network provisioning
• Own hardware procurement for the lab: work with teams on what's coming, order ahead of need, chase quotes and lead times, keep receiving and inventory in order
• Own RMA processes end-to-end for lab hardware: triage, swap, ship, track, close; drive vendor escalations
• Build one-off setups engineers ask for: debug and programming harnesses (JTAG/SWD/UART), remote power switching, console and KVM access, and improvised fixes for hardware running outside designed conditions
• Oversee smarthands and contractor work on-site (electricians, cabling crews, riggers, integrators, OEM field engineers): scope it, dispatch it, validate quality
• Handle physical access, asset auditing, and lifecycle tracking for hardware that is expensive and sometimes pre-release
• Build playbooks and lightweight tooling (inventory, remote power, console access) so remote teams can use the lab without pinging you for every step

• Hands-on experience running a hardware, systems, or datacenter lab or similar physical compute environment: rack-and-stack, copper and fiber cabling, power distribution (PDUs, circuits, phase balancing)
• Working knowledge of server hardware and its management plane: BMC/IPMI/Redfish, BIOS/UEFI configuration, serial console, PXE/network boot, firmware updates
• Enough Linux and networking (VLANs, DHCP, DNS, basic switch config) to take a bare machine to a state a remote engineer can use
• Have managed vendors, contractors, quotes, and freight for physical equipment
• Comfortable and safe with power tools, bench power supplies, and a wiring or rack elevation diagram
• Work independently with minimal supervision, know when to ask, bias toward action, and communicate clearly when flagging risks or blockers

• Stood up a lab or datacenter room from empty: layout, power and cooling budget, first round of purchasing
• Time around high-density accelerator systems (GPU/TPU-class racks), liquid cooling, or pre-release/NPI hardware bring-up
• Electronics bench skills: soldering and rework, scope and multimeter, custom cable and debug harnesses, JTAG probes, light fabrication (3D printing, drilling, mounting)
• Supported hardware security, firmware, or platform bring-up work
• Experience with DCIM or asset-management tooling (NetBox, ServiceNow, etc.)
• Bought hard-to-source or specialty hardware and dealt with the compliance and logistics that come with it

This role is based in Anthropic's San Francisco office (500 Howard St). Travel to nearby lab spaces is required, so expect to be in-office more frequently than Anthropic's standard 25% hybrid baseline. The annual salary range is $320,000—$405,000 USD.