Salesforce this morning announced that starting in June, customers who subscribe to its Einstein Vision & Language services will gain access to named entity recognition. Concretely, this means that sales reps will be able to tap the recently launched Einstein Voice Assistant and Einstein Bots to update fields in customer relationship management forms — for instance, if a rep says, “Update the deal size to $500,000,” natural entity recognition will understand that “$500,000” is a “Money” entity and change the corresponding “Deal Amount” field to “$500,000.”
As Zineb Laraki, Salesforce product lead for Einstein Vision & Language, explained in a blog post, named entity recognition is similar to word classification in that it allows users to identify words in pre-defined categories (i.e. name, organization, date, phone number, and website address). If the forthcoming version of the Einstein Bots platform came across the sentence “Please change my shipping address to 9 Cloud St, Cumulonimbus, CA 90260,” for instance, named entity recognition would enable it to understand that “9 Cloud St, Cumulonimbus, CA 90260” is a “Location” entity and to update the “Delivery Address” field accordingly.
It’s worth noting that most major cloud-hosted natural language services already support named entity recognition. Amazon Comprehend, Amazon’s service that leverages AI to glean insights from and spot relationships in text, enables customers to train AI models on up to 12 custom entities at once. For its part, Microsoft offers an entity recognition skill within its Cognitive Services suite that can extract items of different types from text, and Google Cloud’s natural language API can handle a range of entity and entity sentiment analysis tasks.
But Laraki notes that Salesforce’s goal isn’t necessarily to create the best natural language processing engine in the world. Rather, the idea is to tailor Einstein Vision & Language to customer relationship management and Salesforce clients’ needs. “This goal is very specific. We’re focused on solving problems in a way that drives value for our customers,” said Laraki, noting that named entity recognition will also be available to Salesforce Mobile and Salesforce Service Cloud. “We are working on adjusting functionality and models so we can provide world-class Einstein solutions to our customers.”
Salesforce unveiled a slew of natural language processing services last year during its annual Dreamforce conference in San Francisco, one of which was a product — Service Cloud Voice — that embeds telephony inside Service Cloud. Alongside it launched a voice app development toolkit dubbed Einstein Voice Skills, as well as a revamped Einstein Voice Assistant, the virtual assistant integrated with every app built on Salesforce Customer 360.