Athena: Unified LLM Interaction Framework

Athena is a lightweight, developer-friendly ORM (Object-Relational Mapping)-inspired library designed specifically for interacting with Large Language Models (LLMs) like OpenAI, Anthropic, and Google PaLM. With its easy-to-use API, Athena streamlines the process of making requests to different LLM providers, managing message context, and handling JSON and text-based responses effortlessly.

Key Features:

• Multi-provider support: Integrates seamlessly with OpenAI (GPT models), Anthropic (Claude models), and Google Gemini (e.g., Chat-Bison).
• Context management: Easily manage conversation context across multiple messages for more coherent interactions.
• Flexible response handling: Supports both text and JSON response formats with advanced filtering and validation for JSON data.
• Custom drivers: Extend functionality by implementing custom drivers for other LLM providers.
• TypeScript-first: Fully typed with TypeScript for robust and predictable development.
• Error handling: Built-in mechanisms to handle API errors and edge cases gracefully.
• Lightweight and fast: Focused on simplicity and performance for modern applications.

Why Choose Athena?

If you're a developer building applications that rely on LLMs, Athena provides a unified abstraction to reduce the complexity of managing multiple APIs. Whether you're developing AI-powered chatbots, personalized assistants, or data-processing pipelines, Athena is the perfect tool for seamless integration.

Use Cases:

• AI Chatbots: Easily manage multi-turn conversations and context.
• Data Processing: Extract structured data (e.g., JSON) from text responses.
• Custom AI Models: Integrate with different providers while maintaining a consistent interface.
• Experimental Projects: Quickly prototype ideas without worrying about API-specific details.

Example:

import { Athena } from "athena-ai";

const config = {
  provider: "openai",
  apiKey: "your-api-key",
};

const ai = new Athena(config);

const response = await ai.createMessage({
  messages: [
    { role: "system", content: "You are a helpful assistant." },
    { role: "user", content: "Tell me about the future of AI." },
  ],
  format: "json",
  example: { prediction: "..." },
});

console.log(response.content);