Founding ML Research Engineer

About the Role

Join an early-stage industrial AI team building a smart-glasses product powered by agentic vision-language models. You will own core parts of the multimodal AI stack, bringing real-time assistance to field workflows in sectors such as data centers, energy, and aerospace.

What You'll Do

• Build and ship production agentic VLM pipelines for visual reasoning across industrial procedures and inspections.

• Create evaluation harnesses, capture failure modes, and develop data-to-fine-tuning loops that improve model quality over time.

• Adapt model orchestration and edge inference to changing connectivity, latency, and hardware constraints.

• Develop real-time voice and video interfaces, including conversational and proactive alert experiences.

• Fine-tune open-source multimodal models using approaches such as SFT and RLHF, and optimize them for on-premise deployments.

What We're Looking For

• Typically 2 to 6 years of experience shipping production VLM or multimodal AI systems, with Python and PyTorch.

• Experience taking vision-language systems from development through deployment, including model orchestration, applied fine-tuning, and evaluation.

• Hands-on experience building multi-step agents with real tool use, rather than only single-turn chat or basic retrieval workflows.

• A track record of shipping AI products for end users and owning product outcomes, ideally in a startup or other fast-paced environment.

• Strong computer vision and vision-language background, with interest in wearables and industrial applications.

• Experience with edge inference or on-premise deployment is valuable, including tools such as vLLM, Triton, TensorRT, ONNX, or quantization.

• A bachelor's or master's degree in computer science, machine learning, or engineering, or equivalent practical experience.

Compensation & Benefits

Base salary: $180,000 to $230,000 annually. Visa sponsorship is available.

Location

On-site in San Francisco, United States.