Open role
Founding ML Research Engineer
$200,000 to $260,000
The company
A seed-stage startup deploying wearable AI co-pilots for frontline workers in data centres, energy and aerospace. It has paying enterprise customers and a growing pipeline of large industrial prospects.
The role
You would be a foundational member of the engineering team, owning the multimodal AI stack from day one. That covers production vision-language pipelines, evaluation, edge inference and real-time voice and video AI running on smart glasses in the field.
What you would do
- Ship a production agentic vision-language pipeline on smart glasses for real industrial workflows
- Build and own the evaluation and data flywheel that improves model quality each release
- Deliver edge inference and graceful fallback across varying connectivity and latency budgets
- Build real-time voice and video interfaces for different types of end user
- Drive multimodal model training, including fine-tuning open models for on-premise deployment
What we are looking for
- Two to six years shipping production vision-language or multimodal AI systems in Python and PyTorch
- Agentic vision systems shipped to real users, with ownership of the model and orchestration layer
- Applied fine-tuning, evaluation design and model orchestration experience
- Multi-step, tool-using agent loops built in practice, with evaluation harnesses
- On-site in San Francisco five days a week
Interview process
- Introductory call
- Take-home assignment
- Live multimodal system design interview
- On-site
Key details
- Salary
- $200,000 to $260,000
- Equity
- 0.25% to 0.75%
- Experience
- 2 to 6 years
- Location
- San Francisco (On-site)
- Stack
- Python, PyTorch, LangChain
- Visa
- Visa transfers considered
- Openings
- 1 opening
- Interview
- 4-stage interview process
- Company name
- Shared in confidence once we have spoken
- Reference
- CR-151