Einstein Language is a natural language processing (NLP) feature that analyzes unstructured text. Companies with a large influx of customer support emails, social media comments or chat messages can use Einstein Language to filter through the text to flag important messages or forward to the right department. Einstein Language is a component of the Salesforce Platform.
Made up of the combination of the Einstein Intent API and the Einstein Sentiment API, Einstein Language uses natural language processing models to identify emotion and motivation behind a customer inquiry in order to better categorize next steps. Custom and pre-built models can be input into the Einstein Language APIs to cut down on manual work of having to read through every customer inquiry prior to taking action. For example, Einstein Language can identify that an email has a troubleshooting question about a particular product and can relay it to the relevant support technician or it can analyze that a customer has a complaint about the website and can alert a customer service agent.
Einstein Language can be integrated into company applications, social media channels and mailboxes. Currently, it can only support English.
Einstein Intent API - categorizes customer messages based on actions. Applying this API to unstructured text gives companies the ability to determine which products are being discussed, which departments are involved and which employees need to respond. This automates the process of getting every inquiry, comment or question into the right employee's inbox.
Einstein Sentiment API - labels customer text based on projected emotion- positive, negative or neutral. This API can be used to collectively analyze leads, successfully address dissatisfied customers and monitor brand perception across multiple channels.