AI · vision-language
ABOUT THE ROLE
Join an early-stage industrial AI team building wearable AI tools for field technicians. You will own key parts of the multimodal AI stack, taking vision-language and agentic systems from development into real-world production.
WHAT YOU'LL DO
- Build and ship agentic vision-language model pipelines for smart glasses, including multi-step visual reasoning and tool use in industrial workflows.
- Develop evaluation frameworks, capture model failures, and build data and fine-tuning loops that improve quality with each release.
- Adapt model connectivity and edge inference for varied network conditions, latency requirements, and hardware constraints.
- Create real-time voice and video interfaces, including conversational and proactive alert experiences.
- Fine-tune and optimize multimodal models for enterprise deployments, including on-premise environments.
WHAT WE'RE LOOKING FOR
- Two to six years of experience shipping production vision-language or multimodal AI systems, with ownership of model and orchestration layers.
- Strong Python and PyTorch skills, plus experience with tools such as vLLM, Triton, Ray Serve, ONNX, TensorRT, Hugging Face Transformers, and Docker.
- Hands-on experience with visual reasoning, multi-step agent loops and tool use, evaluation design, data flywheels, and applied fine-tuning such as SFT or RLHF.
- Product engineering experience delivering AI features to end users, ideally in a startup or high-ownership engineering environment.
- Background or demonstrated interest in computer vision, wearables, industrial