ML Research Intern

You’ll help turn research ideas into production-grade models. We’re building computer vision systems that understand engineering drawings at a level no existing tool can — and you’ll be part of making that happen.

• Research and prototype ML approaches for clash detection and drawing analysis
• Build and iterate on CV pipelines for engineering documents
• Design model evaluation frameworks and benchmark datasets
• Work with large-scale proprietary PDF and BIM datasets
• Experiment with multimodal LLM approaches for structured data extraction
• Optimize models for inference speed and production reliability
• Document experiments and contribute to internal knowledge base

• Strong ML foundations (loss functions, training dynamics, regularization)
• Proficient in Python and PyTorch (or JAX / Tensor Flow)
• Familiarity with core CV concepts (CNNs, transformers, object detection)
• Ability to read and implement ideas from research papers
• Bonus: OCR, document understanding, or multimodal models

• YC-backed — work at a well-funded startup at the forefront of AI
• Proprietary datasets — access to AEC datasets that don’t exist anywhere else
• Oxford researchers — mentorship from technically excellent founders
• Your models go to production — real impact, not a shelf project

Excited by this? Send us a short intro, any papers you’ve read or experimented with, and whatever you’ve built in ML.