Salesforce Service Cloud offers a multichannel Bot that integrates with Salesforce data. Salesforce Einstein and the power of Salesforce data makes the Bot smarter to complement the support chat experience and helps the team to do more in less time. This article explains How to build Einstein Bots in Salesforce.
- Dialogs
- Dialogs are conversation snippets guiding bot interactions.
- Each dialog contains a dialog intent for optional NLU training.
- The bot transitions between various dialogs in a conversation.
- Dialogs manage specific aspects like Welcome, Main Menu, Order Status, Location and Hours, Transfer to Agent.
- Different dialogs handle distinct portions of the overall conversation flow.
- Dialogs facilitate the organized progression of a customer’s interaction with the bot.
- The first dialog you present to your customer is the Welcome dialog.
- Main Menu dialog present the main capabilities of your bot to your customers. The menu items that you add when you create a bot are saved in the Main Menu dialog.
- Dialog intents
- Dialog intents represent customers’ reasons for engaging with the bot.
- Intents are actions and typically commence with verbs like buy, return, find, look up, cancel, and edit.
- Dialogs can incorporate dialog intents optionally.
- Training the bot with dialog intents establishes a learning model for understanding customer actions.
- Intents are especially beneficial when customers interact via typed messages in the chat window.
- For understanding customer requests, intents play a crucial role in guiding the bot’s responses.
- Entities
- Entities are data types gathered from customers during interactions.
- Salesforce offers system entities, including Text, Date Time, Date, Money, Number, Person, Location, Organization, Percent, Boolean, and Object (standard/custom).
- Custom entities can be created based on specific data requirements.
- Entities aid in collecting and managing diverse types of information.
- Salesforce integrates various standard and customizable entities for versatile data handling.
- Variables
- Variables serve as containers for storing specific customer-collected data.
- Variables are utilized in dialog actions, functioning as both inputs and outputs.
- Variables can be seamlessly inserted as text in messages, enhancing the dynamic nature of interactions.
- These data containers play a key role in managing and utilizing customer-provided information within the conversation flow.
- Utterances
- Include the common conversational phrases used by the customers to describe an intent.
- Serves as the building blocks of intent models.
- A single intent contain many utterances because there are many ways to describe the same intent.
Enable Bots
- Navigate to Setup and type “Einstein Bots” in the Quick Find box, then choose “Einstein Bots”.
- Toggle the switch next to Einstein Bots to enable the feature.
- Navigate to “My Bots” and initiate the creation of a new bot.
- Select the ‘bot type’
- Select the template form the available templates for your bot type.
- Personalize your bot.
- Choose the Main menu items.
- Link your bot to an existing deployment in the next step.
- Each bot must have a unique embedded service deployment. (Creating a chat queue and deployment will be covered in another article).
- Kick off the intent model.
You will be redirected to the Einstein Bot builder page.
To learn more click, First, What’s a Bot? (salesforce.com)
Take Five Consulting is a technology company, based in Virginia U.S., that specializes in the Mortgage Banking vertical especially LOS implementation and application development. Take Five Consulting creates and implement mortgage technology and software specifically for Mortgage Industry.
