San Francisco startup building wearable AI co-pilots for industrial field workers on smart glasses.
Founding AI Engineer
You'll spend mornings debugging vision-language model pipelines that run on-device, then afternoons whiteboarding with field engineers about what actually breaks at a substation or inside a data center. One day you're distilling a 70B model to run with sub-watt thermal budgets; the next you're in a hard hat watching your agent guide a technician through a valve inspection. This is hands-on work with immediate, physical stakes—your model doesn't just score on a leaderboard, it keeps someone from wrong-heading a critical repair.
What they're looking for
- 1-5 years shipping ML systems in PyTorch, with at least one project deployed to production or real hardware.
- Solid grasp of multimodal architectures—vision-language models, edge optimization, or agentic reasoning loops.
- Python fluency and comfort with Langchain or comparable orchestration; Docker for reproducible environments.
- Willingness to travel occasionally to industrial sites for field testing and direct user feedback.
- Genuine interest in industrial applications: energy infrastructure, aerospace maintenance, or similar physical-world domains.