Onedot Platform News

It is now possible to automatically use feedback from several solved feedback tasks during an onboarding session.

This allows, among other things, to consider feedback already given without waiting for the remaining feedback, as well as to distribute feedback tasks appropriately within the team.

The platform module for automatic product classification with the help of artificial intelligence (AI) has been substantially expanded once again:

New additions include additional algorithms, improved tokenization, smarter entity recognition, and word embedding models that have been trained on the Onedot Product Graph, which already contains over 100 million products, to achieve up to 200% higher accuracy than before, depending on data quality.

The dialog for selecting files in the Onedot app now also has a built-in filter, so that deep hierarchies with a lot of data can be browsed quickly.

Data records with many columns, as is often the case with complex catalogs with many product attributes, are now displayed much more smoothly and can be processed more quickly. The underlying algorithms have been optimized to better calculate possible changes.

Setting up a customer-specific client with its own domain, such as customer.onedot.com, has been simplified.

In addition to the possibility for white labelling, i.e. customer-specific presentation, colouring, etc., Federated Login via existing customer logins and country-specific data storage is also possible with a separate client.

The placement of the short info that is displayed to show larger text attributes has been optimized.

In addition, the initial delay has been slightly increased. Both measures have the effect that the short info gets in the way of the user in fewer cases, but is still reliably displayed when needed.

The action Undo All in the Onedot app has been renamed to Discard Changes to clarify the behavior of this action, a tooltip is now also available that describes the action.

Discard Changes resets the status of a feedback task to the last saved state.