AI Tools

Train the damage-detection model and extract structured data from vehicle photos.

Damage Recognition

Visual annotation

Draw bounding boxes around dents, scratches, and damage on auction photos to train the damage-detection model. Each annotation feeds the next training run.

YOLO pipeline Active learning loop

Text Recognition

OCR extraction

Extract VINs, odometer readings, compliance dates, and number plates from photos via OCR. Review predictions, correct mistakes, and flag unreadable data.

VIN / odo / rego Correction queue

ML Photo Labeler

GPT-4o training

Select listing photos and send them to GPT-4o Vision for automatic damage labeling. Labels are converted to YOLO format to train your local models.

GPT-4o Vision Auto-label pipeline

All tools contribute to the same proprietary model — every annotation and label improves auto-detection on the next scrape.