$f$(👷‍♂️)

IMPORTANT

This package has been deprecated. Please move to our new package - aifn!

From weco to aifn

When migrating, here's what you need to know:

Package

• weco: Install using

  pip install weco

• aifn: Install using

  pip install aifn

---

Building AI Functions

• weco: Use the build method to create a function:

  from weco import build
  function_name: str, function_version: int, function_description: str = build("identify objects in an image")

• aifn: Use the build method to create a function with type AIFunction:

  from aifn import build
  object_identifier: AIFunction = build("identify objects in an image")

---

Calling AI Functions

• weco: Query an AI function using the query method:

  from weco import query
  response = query(
      fn_name="BusinessIdeaAnalyzer-XYZ123",
      text_input="A subscription service for personalized, AI-generated bedtime stories for children."
  )
  output = response["output"]

• aifn: Query an AI function by creating an AIFunction object:

  from aifn import AIFunction
  idea_analyzer = AIFunction("BusinessIdeaAnalyzer-XYZ123")
  output = idea_analyzer("A subscription service for personalized, AI-generated bedtime stories for children.").output

A client facing API for interacting with the Weco AI function builder service!

Use this API to build complex systems fast. We lower the barrier of entry to software engineer, data science and machine learning by providing an interface to prototype difficult solutions quickly in just a few lines of code.

What We Offer

• The build function enables quick and easy prototyping of new functions that use foundational models through just natural language. We encourage users to do this through our web console for maximum control and ease of use, however, you can also do this through our API as shown here.
• The query function allows you to test your newly created functions and deploy it in your code.

We provide both services in two ways:

• weco.WecoAI client to be used when you want to maintain the same client service across a portion of code. This is better for dense service usage or in an object oriented paradigm.
• weco.query and weco.build to be used when you only require sparse usage or a functional paradigm.

Features

• Structured Output
• Grounding (Web Access)
• Multimodal (Language & Vision)
• Versatile Client (Synchronous, Asynchronous, Batch Processing)
• Interpretable (Observe Reasoning Behind Outputs)

Getting Started

Install the weco package simply by calling this in your terminal of choice:

pip install weco

When using the Weco API, you will need to set the API key: You can find/setup your API key here. Once you have your API key, pass it directly to the client using the api_key argument or set it as an environment variable as shown:

export WECO_API_KEY=<YOUR_WECO_API_KEY>

Example

We created a function on the web console for the following task:

"Analyze a business idea and provide a structured evaluation. Output a JSON with 'viability_score' (0-100), 'strengths' (list), 'weaknesses' (list), and 'next_steps' (list)."

Here's how you can query this function anywhere in your code!

from weco import query
response = query(
    fn_name="BusinessIdeaAnalyzer-XYZ123",  # Replace with your actual function name
    text_input="A subscription service for personalized, AI-generated bedtime stories for children."
)

For more examples and an advanced user guide, check out our function builder cookbook.

Happy building $f$(👷‍♂️)!

Contributing

We value your contributions! If you believe you can help to improve our package enabling people to build AI with AI, please contribute!

Use the following steps as a guideline to help you make contributions:

1. Download and install package from source:

   git clone https://github.com/WecoAI/weco-python.git
   cd weco-python
   pip install -e ".[dev,docs]"

2. Create a new branch for your feature or bugfix:

   git checkout -b feature/your-feature-name

3. Make your changes and run tests to ensure everything is working:

   Tests can be expensive to run as they make LLM requests with the API key being used so it is the developers best interests to write small and simple tests that adds coverage for a large portion of the package.

   pytest -n auto tests

   If you're just making changes to the docs, feel free to skip this step.

4. Commit and push your changes, then open a PR for us to view 😁

Please ensure your code follows our style guidelines (Numpy docstrings) and includes appropriate tests. We appreciate your contributions!