Overview

A great example of how you can use GenAI to answer questions that are mostly the same, but slightly different. Take the recurring processes where a list of questions needs to be answered. RFPs (Request for Proposals) are questionnaires which are usually sent out to a service provider to assess their qualification for a project. For the service provider, although each customer asks the questions in a unique way, the content is similar across potential customers. The RFP Starter app represents a smart and time-saving alternative to manually finding similar responses each time and purely relying on the knowledge management by the respective teams by providing the following GenAI-powered capabilities:

• Historical RFPs and other static knowledge form the foundation of a knowledge base
• The knowledge base is being used to generate responses for uploaded RFP documents
• Keep the human in the loop because the user can still edit generated responses (with AI assistance to keep leveraging the information in the Knowledge Base)
• Create a self-learning system by continuously updating the Knowledge Base with the redacted responses
• Enable the rest of the organization to ask questions and tap into the potential of this shared knowledge base through the 'Quick Answer' section.
• LLM Provider Flexibility: Explore the LLM capabilities of Mendix Cloud GenAI, (Azure) OpenAI and Amazon Bedrock or add your own LLM connector to utilize the AI model that best suits your requirements.

Documentation

Configuration

All modules that are needed to interact with an LLM using and configuring Mendix Cloud AI resources are already installed. The app further contains demo data including reference RFPs (Requests for Proposal).

First-time Setup

In order to setup the starter app for the first time, make sure the encryption key is set before anything else and then start the app. You need to log in as an administrator. For a quick demo setup using Mendix cloud AI resources, the following three actions need to happen in the listed order:

1. Connect the LLM

To use Mendix cloud AI resources, configure access in runtime by importing a resource key for a knowledge base resource.

2. Assistant infrastructure

Before responses can be generated, an assistant configuration is required. Make sure to give it a recognizable name, for example 'RFPAssistant' and press save.

3. Populate knowledge base 

This is a manual initialization action that must be executed to populate the knowledge base. That way the knowledge base is in sync with the Mendix app data and the assistant can search through reference RFPs and other data within the knowledgebase. Simply press the link button and wait until it is finished.

Admin tools

Demo data

There is the option to experience this starter app using demo data. Demo data is loaded by default in the after startup microflow, and after completing the first-time setup steps the knowledge base will automatically be in sync.

Need to reset everything to the initial state? Simply press all four buttons after each other in the demo data section and you are good to go!

Masterdata

Manage

- Product lines

- Quick answer FAQs

- Tone of voice.

Reference documents

- Upload RFPs (or other data) for reference in the 'Manage documents' section.

Customize RFP Assistant

This app serves as a starting point and there are many ways to customize your app:

• Add your custom styling (see Customize Styling for more details)
• Use any other type questionnaire than just RFPs 
• Customize product lines
• Redesign the overviews and chat page to your needs.
• Add custom pre- or post-processing logic into the action microflow that interacts with the LLM
• Leverage in-app and external data by combining Function calling, RAG and Knowledge base interaction.
• Adapt the system prompt and add/change available functions to leverage in-app data and perform actions on behalf of the user using the assistant.
• Bring your own knowledge base.
• Add your own LLM provider connector.