Our customers are using Bot Core with multiple bots that we have deployed. Our Intranet Bot solution, MeshBOT is also built using the same. Bot Core is built as a flexible service that can leverage any AI service.
The knowledge base is the primary source of knowledge for the bot. Using this the bot can provide responses to all supported questions or operations that the user requests from the bot.
KB has support for the following:
Handling Singular Q&A responses and Contextual responses.
Creation of Guided Conversations
Guided Conversations – are like workflows of different responses and questions that the bot asks a user to guide them to the desired outcome or final conversation state.
Entity Fulfillment & Actions
Ability of the bot to ask relevant questions to capture all required entities to perform an action.
The action that a bot performs based on the user question. Action can be calling a webservice or executing a SQL stored procedure etc.
Message Definition Language (MDL) – MDL is a HTML-like scripting language using which you can define the bot responses. Once a message is defined using MDL, BotCore ensures that the same will be rendered on various supported Bot channels without any changes. MDL supports text, images, hyperlinks, buttons, cards and carousels.
Bot Memory & State – In order for the Bot to remember the context of a conversation, memory and state management is handled within the BotCore. All you have to do is to define the intents, entities & relationships and context handling happens automatically within the bot.
Standard Responses – Standard set of responses of the bot can be easily configured
Security module provides the authentication and authorization mechanisms for the bot.
Using the trigger module, the bot can initiate a conversation with a user based on certain business rules or triggers
Just like any other software, bits also need to be administered and maintained
Training Module – With this module, an admin can train the bot to learn additional intents & entities.
Maintenance Mode – Admin can put the bot in maintenance mode to test the results of any new training or actions.
Error log – log of bot core errors that can help in debugging production issues.
Chat logs – Bot logs each conversation that it has had with any user.
Zero Intent log – Bot will log anything that it didn’t understand or was not found in its knowledge base.
User and Channel management