Einstein Photo Tagger
USPTO-Patented ML Photo Auto-Tagging
The Challenge
Professional photographers shoot thousands of images per session. Every image needs metadata — keywords, categories, scene descriptions — to be searchable and sellable. The standard workflow is manual tagging, which is tedious, inconsistent, and the first thing skipped under deadline pressure.
What I Built
A Lightroom plugin that connects to Salesforce Einstein Vision API for automatic image classification:
- Import Hook — Plugin integrates into existing Lightroom import workflow
- ML Classification — Images analyzed for scene detection, object recognition, and category prediction
- Metadata Mapping — ML predictions translated to industry-standard IPTC/XMP fields
- Feedback Loop — Photographer corrections improve future predictions
Tech Stack: Adobe Lightroom SDK, Salesforce Einstein Vision API, IPTC/XMP metadata standards
The Result
USPTO patent granted (US20210241019A1) for “Machine Learning Photographic Metadata.” Co-invented with Paulson McIntyre at Salesforce. Built years before AI went mainstream — demonstrating production ML integration in a creative industry workflow.