Research Scientist, Computer Vision
Core
Advancing computer vision, deep learning, and NLP to power automated takeoff and estimation for construction drawings.
Role type
Research Scientist (Computer Vision & Deep Learning)
Builds
Multimodal systems combining visual layout, geometry, and text for automated takeoff and estimation.
Domain
Construction industry + Computer Vision / Deep Learning
Deliverable
production ML models
Required skills
object detection, image segmentation, OCR and layout analysis, self-supervised learning, multimodal/vision-language models, neural network training at scale, Python, PyTorch, GPU/distributed training, data judgment
Preferred skills
vector graphics and CAD formats (SVG, DXF, DWG, IFC), LLM and NLP experience, model optimization (quantization, pruning, knowledge distillation), cloud environments (GCP, AWS, Azure)
Technologies
PyTorch, Python, GCP, AWS, Azure
Responsibilities
Own research problems end to end (framing, data, experiments, ablations, shipping models), design evaluation frameworks, stay current with CV/ML research and evaluate relevance to construction problems
Seniority
Mid-Senior (2–8 years experience)
