Mendix Intelligent Business Assistant chatbot

Content Type: Module
Categories: Utility,Data,Artificial Intelligence

Overview

Explore your enterprise data with this GenAI-powered Mendix Business Assistant chatbot. Traditional querying methods can be rigid, time-consuming, and costly, often requiring SQL expertise that most business users lack. This innovative chatbot addresses these challenges by providing a natural language interface to your enterprise data. Simply ask questions in plain English and instantly receive answers and insights.

 

By combining the flexibility of Generative AI with the speed and structure of low-code development, the Mendix Business Assistant delivers an intuitive, conversational experience that empowers real-time decision support. Leverage this powerful tool to enhance productivity, streamline operations, and drive innovation within your organization.

 

Key Features:

  • Natural Language Querying: Ask business-related questions in plain English (e.g., “How many appointments were scheduled this week?”), and get responses based on live data—no SQL skills required.
  • Text-to-SQL Generation: Translates user intent into backend SQL queries, queries the Mendix database, and delivers human-readable responses.
  • GenAI Chatbot Companion: A conversational interface that mimics human interaction and provides contextual assistance to end users.
  • Agentic Task Automation: Functions as an AI agent that understands intent and executes workflows—scheduling, retrieving insights, or triggering actions—directly from user input
  • Contextual Personalization: Adjusts responses and actions based on user role (e.g., service requester, service provider, or administrator), enhancing relevance and usability.
  • Plug-and-Play Flexibility: Can be adapted for any domain or Q&A scenario where structured data access is needed—finance, HR, supply chain, customer service, or operations.

Documentation

https://youtu.be/ynNhFzAOIkM

The Business Assistant can directly be embedded within the application, once associated with the right data entities using the configuration steps called out, it can seamlessly function as the chatbot for that particular application accessing the data sources to provide the necessary information for the users.

 

How to Use – Configuration

1. Login as Admin - Access the application with admin credentials to set up the required configurations.

 

2. Configure Azure OpenAI Endpoints  

  • Navigate to Deployed Model
  • Enter and save your Azure OpenAI endpoint details (model, key, and endpoint URL) to enable AI-powered interactions.
  • If Azure OpenAI is not the subscription of your choice, similar configuration needs to be done for other subscriptions (for ex. Amazon Bedrock)

 

3. Configure Postgres Vector Database

  • Navigate to PgVectorDB.
  • Provide connection details for your Postgres Vector Database, which will be used to store and retrieve embeddings.

 

4. Configure the prompt to be used for chatbot

  • Create search prompt in the format shared -

You are an AI assistant that generates valid SQL queries based on user descriptions. Always return a syntactically correct SQL query for the {dialect} specified, based on the given database schema and user question.

Response Format:{  "output": ""}

Use the latest `question`. If it lacks enough detail, refer to `previous_question` and `previous_assistant_response`.

**Database Schema: (replace with your Application schema in JSON format). Sample provided below'{  "Tablename":

[  {    "Text-column": "string",    "Date-column": "2025-04-29T01:44:31.090Z",    "Number-column": 0,

}]

…Additional tables

}”

 

5. Upload Knowledge Base Content

  • Navigate to KnowledgeBase.
  • Upload relevant user guide or documentation files that you want the AI to reference.
  • Click the Embeddings button.
  • Repeat this embedding step each time you add or update content in the KnowledgeBase to keep your AI responses accurate.

Please note that the application database must be an external database, not an internal or built-in one. Upon completion of the above steps, the business assistant will be fully operational, enabling your user base to submit queries and receive responses from the database in a clear, human-readable format.

 

The Mendix business assistant is a flexible, industry-agnostic solution. It includes example use cases in healthcare, manufacturing, and retail, and can be applied to additional industries or further usage scenarios within these sectors.

 

Healthcare: Patient appointments

Manufacturing: Equipment Maintenance 

Retail: Order Support Handling

 

Releases