Machine Learning Engineer - Multimodal

As a Senior Machine Learning Engineer, you will own the ML lifecycle for the vision-language models that unify our perception and reasoning capabilities, building systems that understand manufacturing drawings and design intent.

What You'll Do:

• Research, develop, and deploy multimodal models that understand content from manufacturing documentation end-to-end, replacing traditional OCR-to-NLP cascades with unified architectures.
• Work alongside the core-engineering team to build and maintain annotation tooling, implement active learning loops, and engineer synthetic data augmentation strategies.
• Develop evaluation frameworks that quantify system behavior and user impact, extending beyond traditional benchmark metrics.
• Collaborate with the other members of the machine learning team to set the technical and product roadmaps for the AI platform.
• Burn down the long tail, as every percentage point of accuracy maps to man-years of time savings at our scale.

What We're Looking For:

• 5-8 years of professional deep learning experience, with at least 2 years working on multimodal models leveraging image and text modalities.
• Strong Python and PyTorch fluency: You've written custom training loops, loss functions, and data loaders from scratch when needed.
• Production deployment ownership: You've shipped models to production and have been responsible for endpoint and model health.
• MS or PhD in Computer Science, Electrical Engineering, or related field preferred; equivalent industry experience valued equally.

Bonus Points:

• You have a passion for manufacturing and believe that the industry needs better software.
• Previously worked in aerospace, defense, or manufacturing, and have experience working with manufacturing data.
• Published research, achieved SOA results on relevant benchmarks, or contribute to open-source frameworks.
• Prior experience working in a high-ownership startup environment.

Compensation: For this role, the target salary range is $160,000 - $250,000 (actual range may vary based on experience).