ChatHive: Conversational AI with GPT-4, Pinecone, LangChain, and MongoDB

ChatHive is a powerful AI-driven chatbot application designed for interactive document-based conversations. Leverage GPT-4 for conversational intelligence, Pinecone for vector-based document retrieval, MongoDB for scalable chat history storage, and a modern frontend built with Next.js, React, and TailwindCSS.

🚀 Deployed Version: (Access restricted to maintain API key security. Follow instructions below to run or deploy your own instance.)

---

Features

• Multi-Topic Conversations: Create and manage multiple topics to chat about.
• Document Uploads: Upload .pdf, .docx, or .txt files for AI-based retrieval and discussions.
• Embedded Storage: Automatically convert files to embeddings and store them in Pinecone namespaces.
• Chat Management: Maintain multiple chats per topic with automatic MongoDB-based history retrieval.
• Integrated OAuth: Google OAuth for secure user login and personalized chat storage.
• In-Browser Management: Create, manage, and delete Pinecone namespaces directly from the browser.

---

Tech Stack

Backend Tools: LangChain, Pinecone, OpenAI GPT-4.
Frontend: Fully responsive design with TailwindCSS.

---

Screenshots

Main Chat Interface

Settings Page

---

Getting Started

Follow the instructions below to set up and run ChatHive locally.

1. Clone the Repository

git clone https://github.com/hisachin/chathive.git --branch master

Pinecone Setup

1. Create an account on Pinecone.
2. Go to Indexes > Create Index:
   • Name: Choose any name.
   • Dimensions: Set to 1536 (required for OpenAI embeddings) or Set to 384 (required for HiggingFace embeddings).
   • Leave other fields as default.
3. Navigate to API Keys and create a new key.
4. Save the API Key and Environment values.

---

MongoDB Setup

1. Create an account on MongoDB.
2. In your dashboard:
   • Click New Project > Create Project.
   • Under Database Deployments, click Build a Database:
     • Choose M0 FREE for a free-tier cluster.
     • Fill out cluster details and click Create.
   • Add your current IP address to the access list.
   • Under Connect > Drivers, copy the MongoDB URI.
3. Replace <password> in the URI with the password you set for your database user.

---

Google OAuth Setup

1. Go to Google Developer Console and create a new project.
2. Under APIs & Services:
   • Go to OAuth Consent Screen > External > Fill out details.
   • Add .../auth.userinfo.email scope.
   • Add yourself as a Test User.
3. Under Credentials:
   • Create OAuth Client ID:
     • Authorized JavaScript Origins: http://localhost:3000
     • Authorized Redirect URIs: http://localhost:3000/api/auth/callback/google
   • Save the Client ID and Client Secret.

---

Install Dependencies

yarn install

Run the App

1. Run the following command to start the development server:

npm run dev

2. Once the local development environment launches, you can access the app at http://localhost:3000.

Using the App

Chat with Files

1. After logging in with Google, go to Settings at the bottom left corner.
2. Upload your files and assign them a namespace. In this context, a namespace is synonymous with the topic of your conversation.
3. You can upload multiple files to different namespaces and maintain multiple conversations about various topics and documents.

Main Chat Area

1. On the home page, you will see your namespace(s) on the left sidebar.
2. Click on a namespace to open it, create a new chat, and start a conversation with the files embedded for that particular namespace.

Troubleshooting

General Errors

1. Ensure your Node.js version is up-to-date by running node -v.
2. If you're encountering issues with a specific file, try converting it to text first or use a different file. It's possible that the file is corrupted, scanned, or requires OCR (Optical Character Recognition) to convert it to text.
3. Double-check your environment variables and make sure they are properly exposed and configured.
4. Ensure you're using the correct versions of LangChain and Pinecone as required by this repository.
5. Verify that your .env file contains valid API keys, environment settings, and index names.

Pinecone Errors

1. Ensure that your Pinecone dashboard environment and index match the values you provided in your .env file.
2. Confirm that the vector dimensions are set to 1536 as required by OpenAI embeddings.
3. If you're on the Starter (free) plan, be aware that Pinecone indexes are deleted after 7 days of inactivity. To prevent this, send an API request to Pinecone to refresh activity before the 7-day period ends.
4. If issues persist, try starting fresh with a new Pinecone project, index, and a cloned repository.

Credits

1. The frontend of this repository is inspired by the GPT-4 & LangChain repository by mayooear.
2. Significant customizations have been made to enhance features and integrations for this specific project.