Integration · Feeds
Any catalogue that can export a file can have good search.
Most e-commerce platforms, ERPs and PIMs can publish a CSV or JSON export. AACSearch fetches it on a schedule, maps fields to the index and removes products that disappear from the feed.
What the feed connector does
CSV with column mapping
Choose which columns become the id, title, text, URL and tags; any other column is kept as a field you can filter and facet on.
JSON with a root path
Point at the array inside the payload (data.items) and map fields the same way.
Incremental runs
When your feed supports a cursor (since, updated_since), only changed records are fetched on each run.
Facets from columns
Brand, category, size or price columns become filters with live counts.
Or push instead of pull
Documents API
Send up to 500 documents per request from your own job; upsert, create, update or emplace.
Signed webhook
Post changes to the index webhook with an HMAC-SHA256 signature over the timestamp and body.
Specifications
- Formats
- CSV (custom delimiter), JSON
- Documents per run
- 500 by default, up to 5,000
- Schedule
- manual, hourly, daily or weekly
- Deletions
- records missing from a full run are removed
- Push alternative
- documents API, 500 per request
- Webhook signature
- HMAC-SHA256 of
timestamp.body
Before you start
- A public https URL for the export (or basic access we can reach)
- A stable unique id column for every product
- Title and description columns; price, brand, category if you want filters
How to start
- 1
Publish the export
Most platforms can schedule a CSV or JSON export to a URL.
- 2
Create a feed connector
Paste the URL, choose CSV or JSON, map the id, title and text columns.
- 3
Run once, then schedule
Check the first import in the console, then switch the connector to hourly or daily.