Deployment Lead — GPU Cluster Field Deployments

Comet Cloud; Boston (MA); On-site; USD 120,000 - 180,000; Full time

Benefits offered by this job

Hands-on ownership

Access to latest NVIDIA hardware

Deployments with travel

Job summary

Comet Cloud is seeking a Deployment Lead for GPU cluster field deployments in the Boston or Dallas area with travel required. You will be the on-site team lead, partnering with our senior deployment engineer and vendor partners to rack, stack, cable, power up, and validate AI/GPU systems and connect them to customer networks.

You will own the install end to end, leading on-site crews, configuring BIOS/BMC, flashing firmware, and ensuring a clean handoff.

Qualifications

• Experience building HPC/AI/GPU clusters with hands-on work.
• Led on-site crews of engineers and vendor techs during deployments.
• Configured BIOS/BMC, flashed firmware, and used Linux during bring-up.

Responsibilities

• Lead on-site for cluster deployments and own the install end to end.
• Rack, stack, cable, power up, and validate GPU clusters on site.
• Connect new infrastructure to enterprise networks (subnets, VLANs, BGP, ASN, VRF, uplinks).
• Collaborate with senior deployment engineer and vendor partners to hand off to customers.

Skills

HPC/AI GPU clusters

On-site engineering leadership

Network integration (subnets, VLANs, B

Linux system administration

Rack, power, and cabling experience

Tools

BIOS/BMC flashing

Job description

Deployment Lead — GPU Cluster Field Deployments (Boston or Dallas, travel required)

We deploy some of the densest AI infrastructure on the planet — GB300/GB200 NVL72 rack-scale systems, B300 HGX systems, Quantum-3 XDR Infiniband, direct-to-chip liquid cooling — and we commit to commissioning dates in writing. We're hiring the person who leads those installs on the floor.

The job: You're the on-site team lead for cluster deployments. You pair with our senior deployment engineer and vendor partners to rack, stack, cable, power up, and validate GPU clusters — then get them connected to the customer's network and handed off clean. You own the install end to end.

• - Built HPC or AI/GPU clusters with your own hands
• - Led on-site crews of engineers, vendor techs, and remote hands
• - Run structured copper and fiber at high density and kept the port map honest
• - Configured BIOS/BMC, flashed firmware, and lived in a Linux shell during bring-up
• - Connected new infrastructure to an enterprise customer's network (subnets, VLANs, BGP, ASN, VRF, uplinks, etc)

What we offer: A lean team with no layer of process between you and the machines, the newest NVIDIA hardware in production anywhere, and installs where your work is measurably, visibly yours.

Boston metro or Northern Virginia | Significant travel during deployments

Comet Compute is an NVIDIA Inception member building private and hybrid GPU clouds to the NVIDIA Cloud Partner reference architecture. Equal opportunity employer.