This AI-driven support chatbot designed to enhance user experience with its ability to dynamically selects the best AI model based on user queries, ensuring that responses are contextually relevant, accurate, and engaging.
The chatbot is equipped with multiple AI personalities, each representing a different model, making it a versatile assistant capable of handling a wide range of tasks:
-
Lani the Llama (LLama-3.1-8b-Instruct): This personality is friendly and informative, specializing in general support and platform-related inquiries. Lani is the go-to personality for providing clear and concise information about Headstarter and other general topics.
-
Byte the Tech Owl (RWKV): A technical expert, Byte the Tech Owl excels at handling complex coding questions, debugging, and multilingual support. When the conversation requires deep technical knowledge or code-related assistance, Byte is ready to step in.
-
Myra the Myth Weaver (MythoMist): The creative spirit of the team, Myra is perfect for storytelling, role-playing scenarios, and any queries that demand a touch of creativity or imagination. Myra makes interactions more engaging and is ideal for users looking to explore creative solutions.
-
Dynamic AI Model Selection: The chatbot intelligently chooses the most appropriate AI model based on the content of the user's query, optimizing response accuracy and relevance. This dynamic selection ensures that the right personality handles the right task, providing a tailored and seamless user experience.
- Multi-Model Support:
- LLama-3.1-8b-Instruct: Handles general support and platform information.
- RWKV: Ideal for technical questions, coding, and multi-lingual tasks.
- MythoMist: Best for creative and storytelling tasks, role-playing scenarios.
- File Upload Capability: Users can upload documents (up to 40KB) that the chatbot can use to provide more informed responses. Note that it only currently accept .txt files.
- Real-Time Streaming Responses: The chatbot streams its responses to provide near-instant feedback, improving user experience.
- Customizable UI: The front-end is built with React.js and Material UI, providing a sleek, modern interface with a dark theme for comfortable usage.
- React.js: A JavaScript library for building user interfaces.
- Next.js: A React framework for server-side rendering and static site generation.
- Material UI: Provides pre-built components and styling, ensuring a consistent, modern look.
- Node.js: Handles server-side logic and API requests.
- Next.js API Routes: Used for handling backend processes like file uploads and model selection.
- Pinecone: A vector database used to index and search embeddings.
- Hugging Face Inference API: Generates embeddings and powers the AI models.
- Clone the repository:
git clone https://github.com/your-username/ai-support-chatbot.git - Navigate to the project directory:
cd ai-support-chatbot - Install dependencies:
npm install - Set up environment variables:
- Create a
.env.localfile in the root directory. - Add your API keys:
PINECONE_API_KEY=your_pinecone_api_key HUGGINGFACE_API_KEY=your_huggingface_api_key OPENROUTER_API_KEY=your_openrouter_api_key
- Create a
- Run the application:
npm run dev - Access the application:
- Open your browser and navigate to
http://localhost:3000.
- Open your browser and navigate to
- Start a conversation: Enter your message in the chat input field and press Enter or click "Send".
- Upload a document: Click on the "Add Docs" button, choose your file (up to 40KB), and upload it to provide additional context for the chatbot.
- Model Information: Each response from the chatbot indicates which AI model was used, giving you insight into how the system is operating.
- **
page.js: Handles the frontend UI, including chat interactions, message formatting, and file uploads. route.js: Manages backend API routes, including file processing, AI model selection, and handling user messages.
Contributions are welcome! Feel free to fork this repository, make your changes, and submit a pull request.