kiss-ai-stack-core 0.1.0a28

KISS AI Stack's RAG builder core

KISS AI Stack - Core

Effortless AI Stack Building

Welcome to the core of the KISS AI Stack! This module helps you build a stack effortlessly using a simple YAML configuration file. Say goodbye to boilerplate code and embrace minimalism with the KISS principle (Keep It Simple, Stupid).

---

Features

• Centralized Stack Management: Manage multiple session-based AI stacks with lifecycle support.
• Minimal Dependencies: Built using simple, vanilla vendor libraries.
• Tool Classification: Configure tools for your stack to handle specific tasks easily.
• Supports RAG and Prompt-Based Models: Choose the model type that suits your needs.
• Thread-Safe: Reliable operation in multi-threaded environments.

---

Installation

Install the core module using pip:

pip install kiss-ai-stack-core

---

Example Configuration

Here’s an example YAML configuration to set up an AI stack with different tools:

stack:
  decision_maker: # Required for tool classification
    name: decision_maker
    role: classify tools for given queries
    kind: prompt  # Choose from 'rag' or 'prompt'
    ai_client:
      provider: openai
      model: gpt-4
      api_key: <your-api-key>

  tools:
    - name: general_queries
      role: process other queries if no suitable tool is found.
      kind: prompt
      ai_client:
        provider: openai
        model: gpt-4
        api_key: <your-api-key>

    - name: document_tool
      role: process documents and provide answers based on them.
      kind: rag  # Retrieval-Augmented Generation
      embeddings: text-embedding-ada-002
      ai_client:
        provider: openai
        model: gpt-4
        api_key: <your-api-key>

  vector_db:
    provider: chroma
    kind: remote # Choose in-memory, storage, or remote options.
    host: 0.0.0.0
    port: 8000
    secure: false

---

Example Python Usage

Use the core module to build and interact with your AI stack:

from kiss_ai_stack import Stacks

async def main():
    try:
        # Initialize a stack in the stack
        await Stacks.bootstrap_stack(stack_id="my_stack", temporary=True)

        # Process a query
        response = await Stacks.generate_answer(stack_id="my_stack", query="What is Retrieval-Augmented Generation?")
        print(response.answer)

    except Exception as ex:
        print(f"An error occurred: {ex}")

# Run the example
import asyncio
asyncio.run(main())

---

How It Works

1. Stack Initialization: Use Stack.bootstrap_stack to initialize stacks with their configuration and resources.
2. Query Processing: Process queries with Stack.generate_answer, leveraging tools and AI clients defined in the YAML configuration.
3. Tool Management: Define tools to handle specific tasks like document processing or query classification.
4. Vector Database: Use the vector_db section to define how document embeddings are stored and retrieved for RAG-based tasks. Currently, only Chroma is supported.

---

Documentation

Key Methods

• bootstrap_stack(stack_id: str, temporary: bool): Initialize a new stack session.
• generate_answer(stack_id: str, query: Union[str, Dict, List]): Process a query and return a response.

Configuration Highlights

• AI Client: Configure the provider, model, and API key for supported services like OpenAI.
• Tools: Define tools such as general-purpose query handlers or document processors.
• Vector Database: Set up in-memory or persistent storage for RAG-based tasks.

---

Contributing

We welcome contributions! Submit pull requests or open issues to improve this stack.

---

License

This project is licensed under the MIT License. See the LICENSE file for details.