Ebay Lister

1. Upload — drag/drop photos into our ai website
2. Review photo grouping — the AI clusters photos into per-item groups automatically. The VA checks the "Group Review" screen and fixes any mis-groupings (drag photos between groups, merge/split groups, reorder/rotate) before moving on.
3. Generate listings — click Create Listings; the AI writes title, description, category, condition, and item specifics for each group.
4. Review every listing before it publishes (this is the core of the job) — each listing lands in a "Ready to Publish" queue with editable fields (SKU, title, price, condition, category, description) and confidence badges per field. Low-confidence fields need a closer look:

• Comp the item to get an accurate price
• Confirm the AI-read SKU sticker number is correct.
• Check flagged prices ("high variance in sold data — verify").
• Fix category errors before publishing (bad categories cause the eBay publish to fail).
• Resolve duplicate-listing warnings: publish as new, add +1 quantity to the existing listing, or go view the existing one.
• Double-check anything flagged as a box/contents mismatch — the system rebuilds the listing around the real item when the box label doesn't match what's photographed, so title/price/category/condition all need re-verification, and leftover box-only photos may need removing.

6. Publish — approve individually or in bulk ("Publish All"); items can also be discarded instead.
7. Check the Daily QA Report — auto-generated each night (also runnable on demand), logged to Google Sheets. Flags "Missed" items (SKU seen but never published) and "Skipped" items (gaps in the SKU sequence — a box/item that was never listed at all). Mark issues resolved once fixed.

Skills needed: attention to detail, comfortable working a review queue in a web dashboard, enough eBay familiarity to judge whether a price/category/condition looks right. No coding required.