InformaBot

Overview

DocAssistant is an intelligent system designed to store and query runbooks or any other knowledge sources, retrieving relevant documents and prompting a chat model with the same query. While initially leveraging a Milvus vector database, the system is designed to be extensible to any database type, making it a versatile tool for efficient document retrieval and knowledge management.

Setup Instructions

1. How to Set Up a Milvus Server

To set up a Milvus server, follow these steps:

1. Allocate Additional Memory to Docker:

   • Milvus requires a minimum of 8GB of available memory. Docker usually allocates only 2GB by default. Increase Docker memory through the Docker desktop settings under Resources.

2. Download Docker Compose Configuration:

   • Create a directory and download the Docker Compose configuration file for Milvus:

     mkdir milvus_compose
     cd milvus_compose
     wget https://github.com/milvus-io/milvus/releases/download/v2.2.8/milvus-standalone-docker-compose.yml -O docker-compose.yml

3. Run Milvus Using Docker Compose:

   • Start Milvus with Docker Compose:

     docker compose up -d

   • Verify that all containers are running:

     docker ps -a

   • Check the Milvus server logs to ensure it's up and running:

     docker logs milvus-standalone

   For detailed instructions, refer to the Milvus Standalone Setup Guide.

2. How to Populate the Database

1. Place all your runbooks and knowledge documents into the knowledge-source folder.

2. Run the populate_db.py script to populate the Milvus database with these documents:

   python populate_db.py

This script will index the documents and store them in your Milvus server.

2. How to Populate the Model

1. Ensure you have a .env file in your project directory with the following environment variables:

   PROJECT_ID=your_project_id
   PROJECT_URL=your_project_url
   IC_API_KEY=your_api_key

2. Run the query.py script to query the database and generate responses using the IBM model:

   python query.py

The script will prompt you to enter your question, then retrieve relevant documents from Milvus, and finally generate a response using the IBM model.