TRADE FAQ
Where should I start: /trade or /trade/upload?
Use /trade as the main entry point.
- Choose
Upload SKU Datawhen you have a smaller, already-clean file that should go straight into preview. - Choose
Upload Large Feedwhen you have a large or messy supplier feed that needs server-side normalization first. - Use
/trade/uploaddirectly only when you already know you want the clean interactive upload wizard.
If server-side feed profiling is unavailable (unconfigured environment or transient failure), /trade/upload will show a warning and fall back to the interactive wizard for all file sizes.
What is preview classification for?
Preview classification lets you validate file quality, description quality, and likely output shape before you commit to a broader upload or downstream workflow.
Can I leave the page while a preview job is running?
Yes. Use the Jobs page to monitor progress. When the preview completes, View Classification takes you back toward the upload review flow.
Why does returning from Jobs sometimes only show a summary instead of the old results?
Full review can only be restored automatically when the browser still has the matching local session context from the original upload flow. Without that browser state, the app shows an informational fallback and you may need to rerun the file.
What should I do when preview results are low-confidence or unclassified?
Treat that as a data-quality and review signal, not as a system failure. Tighten product descriptions, verify attributes, and escalate rows that need human review before relying on the output.
Does TRADE replace legal classification responsibility?
No. TRADE is decision-support and workflow infrastructure. Final classification, filing, and legal responsibility remain with the customer.
Where do I review SIMA outcomes?
Use the TRADE SIMA surfaces after upload or review flows identify items that need SIMA evidence and workflow action.
How do I get developer or integration help?
Use the developer docs at /docs/developers for API contracts, auth, integration keys, and async job patterns.