intelisys 0.2.7

Intelligence/AI services for the Lifsys Enterprise

Intelisys

Intelisys is a Python library that provides intelligence/AI services for the Lifsys Enterprise. It offers a unified interface to interact with various AI models and services, including OpenAI, Anthropic, Google, and more.

Installation

You can install Intelisys using pip:

pip install intelisys

This will install the latest stable version from PyPI.

For the latest development version, you can install directly from GitHub:

pip install git+https://github.com/lifsys/intelisys.git

Note: If you encounter any issues during installation, make sure you have the latest version of pip:

pip install --upgrade pip

Then try the installation again.

Requirements

• Python 3.7 or higher
• A 1Password Connect server (for API key management)
• Environment variables:
  • OP_CONNECT_TOKEN: Your 1Password Connect token
  • OP_CONNECT_HOST: The URL of your 1Password Connect server

Note: If no local 1Password Connect server is available, the library will fail to retrieve API keys.

Features

• Support for multiple AI models (OpenAI, Anthropic, Google, TogetherAI, Groq, MistralAI)
• Secure API key management using 1Password Connect
• JSON formatting and template rendering
• Asynchronous assistant interactions
• Template-based API calls
• Improved error handling and logging
• Consistent versioning across all package files

Usage

Here's a quick example of how to use Intelisys:

from intelisys import get_completion_api, Intelisys

# Make sure OP_CONNECT_TOKEN and OP_CONNECT_HOST are set in your environment

# Using get_completion_api
response = get_completion_api("Hello, how are you?", "gpt-4")
print(response)

# Using Intelisys class
intelisys = Intelisys(name="MyAssistant", provider="openai", model="gpt-4")
response = intelisys.chat("Tell me about artificial intelligence")
print(response)

Advanced Usage

from intelisys import template_api_json, get_assistant, fix_json

# Using a template for API calls
render_data = {"user_name": "Alice"}
system_message = "You are a helpful assistant. Greet {{user_name}}."
response = template_api_json("gpt-4", render_data, system_message, "friendly_assistant")
print(response)

# Using an OpenAI assistant
assistant_id = "your_assistant_id"
reference = "What's the weather like today?"
responses = get_assistant(reference, assistant_id)
for response in responses:
    print(response)

# Fixing malformed JSON
malformed_json = "{'key': 'value', 'nested': {'a':1, 'b': 2,}}"
fixed_json = fix_json(malformed_json)
print(fixed_json)

Supported Models

Intelisys supports a variety of AI models:

• OpenAI: gpt-4o-mini, gpt-4, gpt-4o
• Anthropic: claude-3.5
• Google: gemini-flash
• TogetherAI: llama-3-70b, llama-3.1-large
• Groq: groq-llama, groq-fast
• MistralAI: mistral-large

New in Version 0.2.7

• Implemented lazy loading for Intelisys class attributes
• Improved performance by initializing attributes only when first accessed
• Introduced properties for model, api_key, and client in the Intelisys class
• Reduced initialization overhead of the Intelisys class
• Updated version number in relevant files
• Maintained all improvements from previous versions

API Reference

For detailed information on available functions and their usage, please refer to the docstrings in the source code.

License

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

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

Changelog

For a detailed list of changes and version history, please refer to the CHANGELOG.md file.

About Lifsys, Inc

Lifsys, Inc is an AI company dedicated to developing solutions for the future. For more information, visit www.lifsys.com.