Edge AI Researcher: On-Device Models that Deliver

Rnb Consultancy; San Francisco (CA); On-site; USD 200,000 - 250,000; Full time

Benefits offered by this job

Relocation support

Immigration support

Job summary

San Francisco AI Lab is hiring to advance on-device agents that run privately on consumer devices. You will own end-to-end model training, from pretraining decisions to distillation and quantization, with a focus on small, efficient models that outperform their size.

You'll work across data mix, objectives, and production runtime, with a clear metric for your stream and checkpoints that reach production. Relocation and immigration support are provided.

Qualifications

• Hands-on depth across the training stack: pretraining, SFT/RLHF/DPO/GRPO and quantization.
• Experience with RL, on-device deployment, or both; systems fluency is a plus.
• Design experiments around metrics that matter and ship checkpoints.
• Experience in frontier labs or startups during high growth.

Responsibilities

• Lead the training of a model family powering on-device agents, including post-training and distillation.
• Own end-to-end capabilities: data mix → objective → evals → production runtime.
• Manage your own workstream with measurable checkpoints sent to production.
• Run experiments and publish actionable write-ups to inform the team.
• Collaborate closely with infra, product, partnerships, and researchers.

Skills

Training stack depth

On-device deployment

Systems fluency (memory, GPU)

Experiment design

Startup/hypergrowth experience

Job description

San Francisco AI Lab is hiring to advance on-device agents that run privately on consumer devices. You will own end-to-end model training, from pretraining decisions to distillation and quantization, with a focus on small, efficient models that outperform their size.

You'll work across data mix, objectives, and production runtime, with a clear metric for your stream and checkpoints that reach production. Relocation and immigration support are provided.