Salesforce Known Issues
Trailblazer Community

Known Issues · Einstein Discovery

Einstein Predictions - Security predicate on supplemental dataset may prevent prediction scores from being evaluated properly

Einstein Discovery

When a model is deployed with a supplemental dataset for field mapping, a security predicate on the supplemental dataset may prevent the Integration User from seeing rows and as a result, the prediction may be evaluated incorrectly and the prediction scores may not be accurate.

Scheduled - Spring '21 ·Reported By 2 ·Updated 8 days ago

Einstein Discovery - Error occurs when accessing a specific Story and Dataset.

Einstein Analytics - Dashboards/Lenses , Einstein Discovery

Sometimes the story becomes inaccessible after refreshing the model. Also, when opening the dataset that is the source of the story, the following error may be displayed. ------------- This page has an error. You might just need to refresh it. [An unexpected error occurred. Please include this...

In Review ·Reported By 1 ·Updated 8 days ago

Bulk Predictions fails due to count query timeout

Einstein Discovery

When choosing a filter for bulk query which turns into a SOQL statement to get the count of rows the result is a time out

Scheduled - Spring '21 ·Reported By 0 ·Updated 15 days ago

Einstein Discovery - Error “This page has an error. You might just need to refresh it. [Not authorized for <datasetId>]"

Einstein Discovery

When a dataset used to create a story is changed, creating a new dataset version, opening the story results in the error “This page has an error. You might just need to refresh it. [Not authorized for <datasetId>]"

In Review ·Reported By 2 ·Updated 22 days ago

Einstein Discovery: RecordOverrides Predictions are throwing an internal server error starting Winter`21.

Einstein Discovery

Due to a known issue in Winter`21 Orgs the RecordOverrides Predictions are throwing an internal server error.

Fixed - Winter '21 Patch 8.0 ·Reported By 1 ·Updated 2020-10-19

Few leads are not being scored using Einstein Discovery due to locking contentions.

Einstein Discovery

When Leads are loaded in bulk using any api, We see few of these leads are not being scored by Einstein Discovery. These failure are happening due to row lock contentions on the lead records. Actual Results: When leads are created few of these leads are not scored by Einstein Discovery Expected...

Fixed - Winter '21 Patch 1.0 ·Reported By 0 ·Updated 2020-10-19

Process Builder for AI Prediction Events sending email errors in Production (Sandbox works fine)

Process Builder , Einstein Discovery

A Known Issue has been identified which can impact Process Builder for AI Prediction Events resulting in many emails getting sent to contacts in a Production org. The process is supposed to pick up the AI score from AI Prediction Events and copy that score to a text field but the check to determine if...

No Fix ·Reported By 1 ·Updated 2020-10-12

Discovery - Prediction Write usage increments for predictions using inactive models

Einstein Discovery

A known issue is causing the "Number of predictions written to Salesforce per day" count to increment for predictions that are using an inactive model. No field values are written, but the count increases regardless.

In Review ·Reported By 1 ·Updated 2020-10-09

Bulk Scoring of data from Model Manager is failing

Einstein Discovery

When attempting to bulk scoring of data from Model Manager it will fail for some records

In Review ·Reported By 0 ·Updated 2020-10-02

Einstein Vision and Language Model Builder Dataset Training Fails with Error: Trainer process failed to complete

Einstein Discovery , Spring 20

Dataset created from CSV File may fail to train with Error: Trainer process failed to complete on Einstein Vision and Language Model Builder Package Version 1.23. Training fails on UI but runs successfully using API endpoint.

In Review ·Reported By 2 ·Updated 2020-08-19

1 to 10 of 32 | Next

Any unreleased services, features, statuses, or dates referenced in this or other public statements are not currently available and may not be delivered on time or at all. Customers who purchase our services should make their purchase decisions based upon features that are currently available.