PromethAI-Backend

Introducing PromethAI

PromethAI is an open-source framework designed to empower users with intelligent AI Agents that assist in navigating decision-making, setting personalized goals, and executing those goals effectively. Built on Python, the project promises an advanced level of AGI (Artificial General Intelligence), capable of adapting and providing recommendations tailored to user goals and preferences.

What is PromethAI?

PromethAI is principally a system that suggests choices based on user-defined goals, recalibrating its advice as per the feedback it receives. It currently specializes in offering recommendations in the food domain; however, the framework is flexible and can be extended to operate in various other sectors.

Key Features of PromethAI

• Autonomous Agents: Designed to function independently, simplifying complex tasks for users.
• User Personalization: Customizes its functionality tailored to the individual user’s needs and preferences.
• Decision-making Support: Utilizes decision trees to guide users through choices, facilitating informed decision-making.
• Asynchronous Operation: The system runs its processes asynchronously, ensuring seamless operations without bottlenecks.
• Application Integration: Developers can explore PromethAI’s app-building capabilities through a separate repository promethAI-GUI.
• Task Automation: Capable of automating tasks and executing decisions without manual input.
• Database Support: Compatible with multiple vector databases through Langchain, enhancing flexibility and scalability.
• Low Latency: Focuses on delivering swift and efficient responses.
• User-Friendly: Designed to be easy to use, ensuring accessibility to users with varying levels of technical expertise.
• Easy Deployment: Streamlined deployment processes ensure users can get started promptly.

Demonstration and Architecture

PromethAI features a demo that highlights its functionalities and user interface, accessible through its official website. The architecture comprises advanced components, delivering robust performance and broad capability.

Setup and Use

To get started with PromethAI:

1. Clone the repository using git clone https://github.com/topoteretes/PromethAI-Backend.git.
2. Navigate to the directory and set up your environment configurations.
3. Ensure your system has Docker and Docker Compose installed. Launch PromethAI using docker-compose up promethai --build.
4. Access the local setup through localhost:3000 to experience PromethAI firsthand.

How PromethAI Works

Each interaction with the AI involves several steps:

1. User queries are vectorized and stored in a Pinecone Vector Database.
2. The AI retrieves relevant past interactions from its memory.
3. It determines the appropriate action based on analysis.
4. Actions and thought processes are stored for future reference.
5. The AI generates a response based on past and current data.
6. This interaction is then archived in the memory for continuous learning.

Available Endpoints

PromethAI comes with a set of APIs to interact programmatically. For example, a POST request can be made to the '/recipe-request' endpoint with a JSON payload to receive a customized response.

Important Notice

PromethAI is in its experimental phase, with no implied guarantees. Users should proceed with caution regarding data security, possible system failures, and operational costs associated with using AI models like GPT-4.

Licensing

The software is distributed under the MIT License, encouraging open and collaborative development.

Acknowledgments

The project draws inspiration from Teenage AGI and Baby AGI projects, contributing to its innovative designs and capabilities.

PromethAI stands as a promising tool for those looking to leverage AI capabilities in everyday decision-making and task execution, demonstrating the vast potential of open-source AGI initiatives.