Member of Technical Staff - Efficient ML

embedding-vc; San Francisco (CA); On-site; USD 100,000 - 150,000; Full time

Job summary

A tech-focused company in San Francisco seeks candidates with expertise in AI simulation development. The role emphasizes optimizing training efficiency, enhancing GPU performance, and ensuring low-latency inference. Applicants should be proficient in methodologies for gradient checkpointing, Nsight profiling, and job management tools like SLURM. The company values in-person collaboration in its dynamic team environment, providing opportunities for innovation and cutting-edge technology implementation.

Responsibilities

• Optimize training efficiency with ML techniques.
• Enhance GPU and kernel performance for AI models.
• Implement inference optimization strategies for low-latency serving.
• Ensure infra reliability with job management tools.

Skills

Dataloaders, fusion, activation remat

Gradient checkpointing

Nsight profiling

Triton/CUDA kernels

Quantization (GPTQ/AWQ)

Tools

SLURM

Kubernetes

Job description

Introducing Moonlake, AI for creating world simulations.

Scope of Work

Training efficiency

• Dataloaders, fusion, activation remat, gradient checkpointing.
• FSDP/ZeRO/tensor+pipeline parallel; NCCL tuning.

GPU + kernel performance

• Nsight profiling, Triton/CUDA kernels, fused ops.
• Flash-attention–style speedups, sequence packing, KV-cache tricks.

Inference optimization

• Low-latency serving, continuous batching, speculative decoding.
• Quantization (GPTQ/AWQ), distillation, pruning.

Infra + reliability

• SLURM/K8s multi-node jobs, checkpoint hygiene.
• Determinism, env pinning, GPU failure handling.

We are committed to being an on-site, in-person team currently based in San Mateo