Eva

Eva is an AI-powered chatbot designed to surpass the functionality of leading GPT Apps, offering a seamless conversational experience. Built with cutting-edge technologies, it aims to provide an open-source alternative for developers and enthusiasts looking to explore AI chatbots.

Table of Contents

• Project Overview
• Features
• Technologies Used
• Installation
• Usage
• Contributing
• License

Project Overview

Eva is an ambitious project that aims to replicate the core functionalities of leading GPT Apps, leveraging advanced AI models and technologies. While the project strives to be a perfect open-source clone of leading GPT Apps, it acknowledges the challenge of replicating the UI design and user experience perfectly. The project is driven by the desire to explore and understand the capabilities of AI chatbots, with a particular focus on semantic services and kernel memory implementation.

Features

• AI-Powered Chatbot: Utilizes advanced AI models for generating contextually relevant responses.
• Semantic Services: Enhances understanding and context in conversations.
• Kernel Memory Implementation: A key feature under development to improve chatbot memory and context retention.
• Localhost LLM Support: Planned enhancement to work with locally hosted language models.

Technologies Used

• Backend: Python FastAPI, Langchain, WebSocket, ScyllaDB, Azure OpenAI Models. OLD [.NET ASP Core Web API, Semantic Kernel, SignalR]
• Frontend: Next.js, TypeScript, Tailwind CSS.

Installation

1. Clone the repository: git clone https://github.com/sp4rkiop/eva-ai.git
2. Navigate to the project directory: cd eva-ai
3. Start the backend development server, needs a ScyllaDB / Cassandra instance.
4. Install dependencies of frontend: npm install or yarn install
5. Start the frontend development server: npm run dev or yarn dev

Usage

To use Eva, simply navigate to the application URL in your browser. You can start a conversation with the chatbot by typing your message in the input field.

Contributing

Contributions are welcome! Please read the contributing guidelines before getting started.

License

Shield:

This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.