IncidentNavigatorProject

An intelligent solution for efficient incident resolution using LLM and RAG.

Table of Contents

1. 📖 Introduction
2. 🛠️ Installation
   • Clone Repository
   • Quick Install/Uninstall
   • Manual Setup
3. 🔧 Launch Server
   • Quick Launch
   • Manual Launch
4. 🔒 Default Admin Credentials
5. ⚙️ Configure LLM
6. 📝 User Evaluation

---

Introduction

IncidentNavigator is a solution designed to streamline recurring incident resolution. Using a Large Language Model (LLM) and Retrieval-Augmented Generation (RAG), it offers:

• Dynamic interaction for clarifying queries.
• Transparent, fact-based responses.
• Adaptability to structured and unstructured data.

Demo: Watch the application in action on YouTube.

Installation

Ensure the following are installed on your system:

• Python 3.12.4+
• Docker

Clone Repository

git clone https://github.com/MrSneaker/IncidentNavigatorProject.git
cd IncidentNavigatorProject

Quick install/uninstall

Install

This script will handle the creation of the virtual environment, installation of required Python packages, and any other necessary setup steps.

sudo ./app/install/install.sh

Note: Dependancies installations could take a while, depending on your network speed.

Uninstall

This script is used to uninstall the application.It performs the necessary steps to remove the application and its associated files from the system.

sudo ./app/install/uninstall.sh

---

Manual Setup

1. Navigate to the installation directory:

   cd app/install

2. Docker services:

   Docker allows you to build and run services along with its dependencies in a containerized environment making it so you do not need to install them locally. As our applications need Weaviate, MongoDB and Redis we thus provide a docker-compose.yml file to make it easier to install and use them.

   • Start :

     docker compose up -d

   • Stop :

     docker compose down

   Note: If you encounter permission issues with Docker on Linux, try running the commands with sudo or ensure your user is added to the Docker group (sudo usermod -aG docker $USER)

3. Set up a Python environment:

   python -m venv .env
   source .env/bin/activate  # Linux/macOS
   pip install -r requirements.txt

4. Populate databases:

   python3 create_dbs.py  # Use [-c | --clear] to reset databases.

---

Launch Server

Quick Launch

sudo ./app/install/start.sh

This script will start all components for you. Navigate to http://localhost:5000 to access the homepage.

sudo ./app/install/stop.sh

This script will stop all running components.

Manual Launch

1. Navigate to the project root directory:

   Ensure you are in the root directory of the project before proceeding with the following steps.

   cd /path/to/IncidentNavigatorProject

2. Set up the Python environment:

   Activate the installed environment:

   source app/install/.env/bin/activate  # Linux/macOS

3. Launch the Flask application:

   python3 app/app.pySimplifiez les phrases pour les rendre plus fluides en français. Par exemple :lhost:5000, you will be directed to the homepage.
   

Default Admin Credentials

The application comes with a default admin account for initial use:

• Email: admin@example.com
• Password: admin

Important: It is highly recommended to change the password after the first login to secure the application.

---

Configure LLM

To configure the Large Language Model (LLM), the admin should follow these steps:

1. Log in to the application using the admin account.

2. Navigate to the Settings page.

3. In the Add New LLM Configuration section, provide the following example details:

   • LLM Model: llama-3.3-70b-versatiile (example model)
   • LLM Model URI: https://api.groq.com/openai/v1/ (example URI)
   • API Key: Provide your API key

4. Save the configuration to enable the new LLM model.

Note: Ensure the API key is kept secure and only shared with authorized personnel.

---

User Evaluation

Participate in our user evaluation by filling out this survey.