WatsonX AutoDeploy

Automated deployment library for IBM WatsonX AI services with LangGraph support.

Overview

WatsonX AutoDeploy simplifies the process of deploying AI services to IBM WatsonX by providing a clean, pythonic interface for:

• Creating custom environments with specified dependencies
• Building software specifications
• Storing and deploying AI services
• Managing the entire deployment lifecycle

Installation

From PyPI (when published)

pip install watsonx-autodeploy

From Source

git clone https://github.com/nicknochnack/watsonx-autodeploy.git
cd watsonx-autodeploy
pip install -e .

Development Installation

git clone https://github.com/nicknochnack/watsonx-autodeploy.git
cd watsonx-autodeploy
pip install -e ".[dev]"

Quick Start

Environment Setup

First, set up your environment variables:

# .env file
WATSONX_URL=your_watsonx_url
WATSONX_APIKEY=your_api_key
WATSONX_SPACEID=your_space_id
MODEL_ID=your_model_id

Basic Usage

import os
from dotenv import load_dotenv
from autodeploy import Deployer

load_dotenv()

def my_ai_service(context, **kwargs):
    # Your AI service implementation
    def generate(context):
        return {"body": {"message": "Hello from WatsonX!"}}
    
    def generate_stream(context):
        pass
        
    return generate, generate_stream

# Deploy with one line
deployer = Deployer()
deployer.autodeploy(my_ai_service)

Advanced Usage

from autodeploy import Deployer

deployer = Deployer()

# Step-by-step deployment with custom parameters
deployer.export_config(
    python_version="3.11",
    channels="conda-forge",
    dependencies=["custom-package==1.0.0"],
    prefix="/opt/anaconda3/envs/custom"
)

deployer.build_environment(
    environment_name="my-custom-env",
    base_runtime="runtime-24.1-py3.11"
)

deployer.build_software_spec(
    spec_name="my-spec",
    spec_description="Custom specification"
)

deployer.store_service(
    deployable_ai_service=my_ai_service,
    service_name="my-service"
)

deployer.deploy_service(deployment_name="production-deployment")

API Reference

Deployer Class

The main class for managing WatsonX deployments.

__init__()

Initializes the deployer with WatsonX credentials from environment variables.

autodeploy(deployable_function)

One-step deployment method that handles the entire deployment pipeline.

Parameters:

• deployable_function: The AI service function to deploy

export_config(python_version=None, channels=None, dependencies=None, prefix=None)

Creates a conda environment configuration file.

Parameters:

• python_version (str, optional): Python version (default: "3.11")
• channels (str, optional): Conda channels (default: "empty")
• dependencies (list, optional): List of pip dependencies
• prefix (str, optional): Environment prefix path

build_environment(python_version=None, environment_name=None, base_runtime=None)

Creates a package extension for the environment.

Parameters:

• python_version (str, optional): Python version
• environment_name (str, optional): Name for the environment
• base_runtime (str, optional): Base runtime specification

build_software_spec(spec_name=None, spec_description=None)

Creates a software specification.

Parameters:

• spec_name (str, optional): Name for the software specification
• spec_description (str, optional): Description of the specification

store_service(deployable_ai_service, service_name=None)

Stores the AI service in WatsonX repository.

Parameters:

• deployable_ai_service: The AI service function
• service_name (str, optional): Name for the service

deploy_service(deployment_name=None)

Deploys the stored AI service.

Parameters:

• deployment_name (str, optional): Name for the deployment

Examples

See the examples/ directory for complete usage examples:

• examples/basic_usage.py - Simple one-line deployment
• examples/advanced_usage.py - Step-by-step deployment with custom parameters

Requirements

• Python 3.11+
• IBM WatsonX AI account and credentials
• Required dependencies (automatically installed):
  • ibm-watsonx-ai>=1.3.34
  • langchain>=0.3.27
  • langchain-ibm>=0.3.15
  • langgraph>=0.6.5
  • python-dotenv>=1.1.1

Development

Running Tests

pytest

Code Formatting

black src/ examples/

Type Checking

mypy src/

Contributing

1. Fork the repository
2. Create a feature branch
3. Make your changes
4. Add tests if applicable
5. Run the test suite
6. Submit a pull request

License

MIT License. See LICENSE file for details.

Support

For issues and questions:

• Open an issue on GitHub

Changelog

v0.1.0

• Initial release
• Basic deployment functionality
• Support for custom environments and dependencies
• One-step autodeploy feature