AI product tagging — how it works
Two-pass Claude vision model auto-tags every post on ingest. Bulk-tag thousands at once with confidence thresholds.
Manual product tagging doesn’t scale beyond a few hundred posts. AI tagging does. Idukki runs a two-pass vision model against every post on ingest:
Pass 1: object detection
Claude Haiku identifies wearable / usable products visible in the image. Each detection gets a bounding box and a confidence score.
Pass 2: catalog matching
Each detection is matched against your product catalogue (title + image embeddings) and the top-3 SKU candidates are returned with confidence scores.
Reviewing suggestions
In Manage Collection → Bulk tagging, you see a grid of thumbnails with the top suggestion overlaid. Click Accept all >80% confidence to bulk-approve in one click.
Still stuck?
Email us at support@idukki.io or open the in-app chat. We answer within one working day.
Talk to a human
