MyChatGPT Demo

Welcome to the MyChatGPT demo repository! This project demonstrates how to build a custom ChatGPT-like AI assistant from scratch using best practices in prompt engineering and development.

https://blog.promptlayer.com/building-chatgpt-from-scratch-the-right-way-ef82e771886e

Overview

In this project, we will:

• Write a System prompt to define the AI's role.
• Build a simple chat interface using Python.
• Connect to logs for debugging and iteration.
• Provide MyChatGPT with contextual information like date and location.
• Add a calculator tool to enhance the AI's capabilities.
• Implement best practices for maintainable development, including production deployments, modular prompting, and regression tests.

Getting Started

1. Clone the Repository:

   git clone https://github.com/MagnivOrg/my-chatgpt-promptlayer.git
   cd my-chatgpt-promptlayer

2. Set Up the Environment: We recommend using poetry for managing dependencies.

   poetry install

3. Configure Environment Variables: Create a .env file to store your API keys.

   OPENAI_API_KEY=your_openai_api_key
   PROMPTLAYER_API_KEY=your_promptlayer_api_key
   

4. Run the Application:

   poetry run python app.py

Features

• System Prompt: Define the AI's role and behavior.
• Chat Interface: Simple CLI for interacting with MyChatGPT.
• Contextual Information: Inject date and location into the prompt.
• Calculator Tool: Use OpenAI's function calling to perform calculations.
• Logging and Debugging: Track and iterate on prompts using PromptLayer.

Screenshots

Example of the System prompt used in MyChatGPT.

Example of a function call to the calculator tool.

Conclusion

This project serves as a foundation for building and customizing your own AI assistant. For a detailed tutorial, please refer to the accompanying blog post.

Happy prompting! 🍰

For more information, visit PromptLayer.