> For the complete documentation index, see [llms.txt](https://docs.inogic.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.inogic.com/whatsapp4dynamics/chatbot.md). # ChatBot **ChatBot** is an **AI-powered conversational assistant** built for Microsoft Dynamics 365 and Dataverse. It helps organizations automate conversations, answer queries using business data, and provide instant support through connected knowledge sources like **Dataverse, SharePoint, websites, uploaded documents, and Azure AI Search.** It uses both embedding models (for understanding and retrieval) and generative AI models (for response generation) to deliver accurate, contextual, and human-like responses. Businesses can configure agents, knowledge sources, and AI behavior to match their internal processes and communication style. ### Setup & Configuration Steps Open the ChatBot setup page.
Click on **Create Azure Resources**. This step provisions required Azure services like search, storage, and supporting infrastructure.
You will be redirected to the Azure portal. Click **Review + Create** to verify all resource settings before deployment.
Click **Create** to start provisioning Azure resources for the chatbot.
Azure begins creating resources. Wait until deployment completes successfully.
### Azure Configuration & Connection After deployment is completed, copy the Azure AI Search service details from Azure.
Paste these search service details into the ChatBot configuration to connect search indexing.
Copy Azure Storage Account details, which will be used for storing chatbot-related files and data.
Paste the storage details into the ChatBot and click **Validate** to verify the connection.
Once validation is successful, the system confirms that Azure services are properly connected.
### Security & AI Model Setup Click on **Manage Security Roles** to define who can access and manage chatbot configurations and agents.
Ensure both required AI models are available: * **Embedding Model** → used for understanding and retrieving relevant information * **Generative AI Model** → used for generating responses
### **Agent Creation** Click **New AI Agent** to create a chatbot instance. Each agent can have its own behavior and knowledge base.
The agent configuration form opens where all settings are defined.
Enter agent name and details. Click **Enhance** to automatically improve or refine the system prompt for better AI responses.
### AI Model Configuration Go to the **AI Models** tab where AI behavior is configured.
Select an existing AI model or create a new one and save it for use.
Choose the **WhatsApp4Dynamics configuration** from the dropdown to link chatbot with CRM context.
Use the preview option to test how the chatbot responds before publishing it live.
### Knowledge Source Setup Click **Add Knowledge Sources** to connect data sources that the chatbot will use for answering questions.
You can choose from multiple source types such as websites, SharePoint, or Dataverse.
### **File Upload Knowledge Source** Click **Upload File** to add documents like PDFs, Word files, or Excel sheets.
Select the required file from your system and upload it into the chatbot.
The system automatically processes the file and creates a searchable knowledge source.
You can remove existing files or add new ones whenever needed.
Upload additional files and click **Upload and Process** so the system indexes the content for AI retrieval.
### **Website Knowledge Source** Choose **Public Website** as a knowledge source.
A configuration pop-up appears to enter website details.
Paste the website URL and click **Add** to include it.
Enable **Include Subpages** if you want the chatbot to also read linked pages, then click **Add to Agent**.
You can later edit, remove, or enhance the description of the website source.
You can add multiple websites and update the agent knowledge anytime.
### **SharePoint Knowledge Source** Choose SharePoint as a knowledge source for internal documents.
Popup will open. Enter a name for this SharePoint source.
Open SharePoint and copy the site URL from the browser.
Choose the relevant chatbot-related SharePoint site. Copy the full site URL from the browser address bar.
Paste the SharePoint URL into chatbot configuration.
Select the appropriate container type for storing and organizing data.
### **Azure App Registration Setup** Go to Azure Portal and open **App Registrations**.
Find the required registered application and open it.
Copy the **Application (Client) ID** from Azure.
Paste it into the chatbot configuration.
Copy the client secret key from Azure.
Paste the key into the chatbot and click **Add to Agent** to connect Azure authentication.
### **Dataverse Knowledge Source** Choose Dataverse as a knowledge source to use CRM data.
Select the required entity and give a meaningful name to the knowledge source.
Choose which entity fields should be used for chatbot search and responses.
Select the field that uniquely identifies customers (like phone).
Confirm selections and click **Add to Agent** to connect Dataverse data.
### **Publishing Agent** Click **Publish** to make the chatbot live and available for users.
System confirms that the agent has been successfully published.
### Agent Management You can edit the agent anytime, which moves it back to draft mode for updates.
All created agents are visible in a grid view where they can be managed, edited, or monitored.