Scikit RAG + OpenAI & Discord App for Defang

This repository contains two projects:

1. Scikit RAG + OpenAI in /app: A Flask-based Retrieval-Augmented Generation (RAG) chatbot using OpenAI's GPT model, scikit-learn, and Sentence Transformers for dynamic knowledge retrieval.

2. Discord App for Defang in /discord-bot: A Discord bot designed for Defang Software Labs, providing helpful resources and interacting with users via slash commands.

---

Scikit RAG + OpenAI

Overview

This application demonstrates how to deploy a Flask-based Retrieval-Augmented Generation (RAG) chatbot using OpenAI's GPT model. The chatbot retrieves relevant documents from a knowledge base using scikit-learn and Sentence Transformers and then generates responses using OpenAI's GPT model.

Prerequisites

1. Download Defang CLI
2. (Optional) If you are using Defang BYOC authenticated with your AWS account
3. (Optional - for local development) Docker CLI

Deploying

1. Open the terminal and type defang login
2. Type defang compose up in the CLI.
3. Your app will be running within a few minutes.

Local Development

1. Clone the repository.

2. Create a .env file in the root directory and set your OpenAI API key, or add the OPENAI_API_KEY to your .zshrc or .bashrc file.

3. Run the command:

   docker compose -f compose.dev.yaml up --build

   This spins up a Docker container for the RAG chatbot.

Configuration

• The knowledge base is the all the markdown files in the Defang docs website. The logic for parsing can be found in ./app/get_knowledge_base.py.
• The file get_knowledge_base.py parses every webpage as specified into paragraphs and writes to knowledge_base.json for the RAG retrieval.
• To obtain your own knowledge base, please feel free to implement your own parsing scheme.
• for local development, please use the compose.dev.yaml file where as for production, please use the compose.yaml.

---

Discord App for Defang

Overview

This is a Discord bot developed for Defang Software Labs. It provides helpful resources in a Discord server and interacts with users via slash commands. The bot is built using Discord's official template.

Features

Slash Commands

/ask: A command to ask Defang-related questions to the bot. The bot accesses the Ask Defang (ask.defang.io) API endpoint for retrieving responses.

/test: A basic command to test functionality using the Discord API, without relying on external APIs.

Development

Project structure

Below is a basic overview of the project structure:

├── .env.       -> .env file (not shown)
├── app.js      -> main entrypoint for app
├── commands.js -> slash command payloads + helpers
├── utils.js    -> utility functions and enums
├── package.json
├── README.md
└── .gitignore

Setup project

Before you start, you'll need to install NodeJS and create a Discord app with the proper permissions:

• applications.commands
• bot (with Send Messages enabled) Configuring the app is covered in detail in the getting started guide.

Install dependencies

cd discord-bot
npm install

Get app credentials

Fetch the credentials from your app's settings and add them to a .env file. You'll need your app ID (DISCORD_APP_ID), bot token (DISCORD_TOKEN), and public key (DISCORD_PUBLIC_KEY). You will also need an ASK_TOKEN to authenticate API calls to the Ask Defang endpoint.

Install slash commands

The commands for the example app are set up in commands.js. All of the commands in the ALL_COMMANDS array at the bottom of commands.js will be installed when you run the register command configured in package.json:

cd discord-bot
npm run register

Running the app locally

After your credentials are added, go ahead and run the app:

cd discord-bot
npm run start

Set up interactivity

The project needs a public endpoint where Discord can send requests. To develop and test locally, you can use something like ngrok to tunnel HTTP traffic.

Install ngrok if you haven't already, then start listening on port 3000 in a separate terminal:

ngrok http 3000

You should see your connection open:

Tunnel Status                 online
Version                       2.0/2.0
Web Interface                 http://127.0.0.1:4040
Forwarding                    https://1234-someurl.ngrok.io -> localhost:3000

Connections                  ttl     opn     rt1     rt5     p50     p90
                              0       0       0.00    0.00    0.00    0.00

Copy the forwarding address that starts with https, in this case https://1234-someurl.ngrok.io, then go to your app's settings.

On the General Information tab, there will be an Interactions Endpoint URL. Paste your ngrok address there, and append /interactions to it (https://1234-someurl.ngrok.io/interactions in the example).

Click Save Changes, and your app should be ready to run 🚀