Deliver Einstein Article Recommendations in Five New Languages (Pilot)
Where: This change applies to Lightning Experience in Enterprise, Performance, and Unlimited editions.
Why: Article recommendations in these newly supported languages work just as they do in English. A single, predictive model controls Einstein’s recommendations for all supported languages in your Salesforce org, and agents see recommendations in the Knowledge component of the Lightning Service Console. Einstein uses natural language processing (NLP) to detect each case’s language, and checks the knowledge article version’s Language field for the article’s language.
Einstein searches for articles based on the case’s language. If none meet the relevance requirements, Einstein next searches for articles in your org’s Knowledge master language, and then in the agent’s default language. This means that agents sometimes see article recommendations on a case in more than one language.
- 100 knowledge articles
- 500 instances of an article attached to a case, also called an “attach”