Tech Lead Manager, GPU Cluster Infrastructure
FAR.AI · Remote · Remote (International) ·
COMPENSATION
$225k – $325k
Listed on the employer's job board (Ashby)
- Tech Lead Manager
- GPU & Kubernetes
- Fully remote possible
- Visa sponsorship
- AI Safety Research
FAR.AI is a non-profit institute dedicated to ensuring advanced AI is safe and beneficial. The Foundations team builds the compute platform and tools that accelerate research, managing a rapidly growing fleet of GPUs across multiple providers.
You will lead the new infrastructure sub-team, balancing technical hands-on work with people management. You will define the technical roadmap for cluster operations, security, and scaling while hiring and mentoring senior engineers.
What you'll do
- Define technical direction and roadmap for the GPU cluster platform
- Design architecture for scheduling, storage, and multi-provider operations
- Hire, mentor, and manage a small team of senior infrastructure engineers
- Establish security protocols for shared clusters running AI agents
- Debug complex failures across node health, GPU, and network layers
- Serve as escalation point for research teams and hardware providers
What you'll need
- Experience leading engineers as a manager, tech lead, or project lead
- 5+ years in systems engineering on production Linux and GPU platforms
- Deep expertise in scheduling, storage, networking, security, or GPU systems
- Production experience with Kubernetes for GPU workloads and batch layers
- Proficiency with Infrastructure as Code tools like Terraform or Ansible
- Strong programming skills in Python, Go, Rust, or C++
Nice to have
- Experience with distributed training infrastructure and NCCL debugging
- Knowledge of distributed storage systems like VAST, Weka, or Ceph
- Expertise in cluster security and sandboxed runtimes
- Background in building greenfield teams and on-call practices
- Experience managing multi-provider platforms with uneven data locality
Stack
- Kubernetes
- Linux
- Python
- Go
- Rust
- C++
- Terraform
- Ansible
- Helm
- ArgoCD
- Prometheus
- PyTorch
- Slurm
How you'll work
- Full-time position, 40 hours per week
- Remote worldwide or hybrid in Berkeley, CA / Singapore
- Working hours should overlap with Berkeley time zone
- No formal on-call rotation yet, but expected to join future rotations
- Visa sponsorship available for in-person employees
What you get
- 94% of health insurance premium paid by organization
- 401(k) plan with up to 2% match
- 25 days paid time off plus 10 days sick leave
- Paid bereavement, family, medical, and pregnancy disability leave
- Work computer and WFH stipend provided
Hiring process
- One hour programming assessment
- Short screening call
- One hour code review interview
- 90 minute systems design and incident response interview
- One hour management and leadership interview
- Paid work trial lasting up to one week