Salesforce announced this morning that as of June, customers subscribed to Einstein Vision & Language services will have access to Named Entity Recognition, a feature that automatically updates fields in Customer Relationship Management (CRM) forms.
This means that vendors will be able to use the recently launched Einstein Voice Assistant and Einstein Bots to update the fields in the forms. For example, if a user says “Update the value of the deal to 100,000 euros”, named entity recognition can recognise that the amount is a “money” entity, and then update the corresponding “amount” field to “€500,000”.
According to Salesforce itself, the technique is similar to word classification, as users can classify words into pre-created categories (e.g., name, organisation, date, phone number, or URL).
Most cloud voice recognition services already support named entity recognition. For example, Amazon Comprehend allows customers to train AI models on up to 12 custom entities at a time. Microsoft offers an entity recognition feature in its Cognitive Services suite that can extract certain items from different types of text. There is also the natural language API from Google Cloud, which can handle a range of these types of tasks.
According to Salesforce, however, it is not necessarily the intention to create the best natural language processing engine in the world. Rather, the idea is to align Einstein Vision & Language with the needs of CRM users (and thus Salesforce customers). “We focus on solving problems in a way that increases value for our customers,” says Zineb Laraki, Salesforce product lead for Einstein Vision & Language. The feature will also be available on Salesforce Mobile and Salesforce Service Cloud. “We are working to customise the functionality and models so that we can provide our customers with world-class Einstein solutions.”