azure-ai-helper

A simple, human-friendly wrapper over Azure AI Vision and Language services.

• Clean API: one-line calls like ai.vision.analyze(path_or_url) and ai.language.summarize(text).
• Human-readable outputs instead of raw JSON.
• Friendly errors with quick hints.

Install

Choose extras for services you need:

pip install "azure-ai-helper[vision,language]"

Or base package only (no SDKs):

pip install azure-ai-helper

Quick Start

from azure_ai_helper import AzureAI

# IMPORTANT: Use your own Vision & Language endpoint/key
# Do NOT reuse credentials from other Azure services
ai = AzureAI(endpoint="https://<your-endpoint>", key="<your-key>")

print(ai.vision.analyze("my_image.jpg"))  # local file
print(ai.vision.analyze("https://example.com/photo.jpg"))  # URL

print(ai.language.summarize("Long text..."))
print(ai.language.sentiment("I love this!"))

Credentials

• You must provide your own Azure Cognitive Services endpoint and key.
• This library never ships with embedded credentials and will not read any demo defaults. You control the endpoint/key.
• Use resource keys for the specific services included here (Vision and Text Analytics). Do not use keys from unrelated services.

Errors

• All errors raise AzureAIError with a helpful hint. Common cases:
  • Missing SDK -> install the corresponding extra.
  • Invalid image URL -> provide a direct .jpg/.png or local path.
  • Summarization unsupported -> we fallback to key phrases; consider enabling the feature in your Azure resource.

Testing

Install dev tools and run tests:

pip install pytest
pytest -q

Roadmap

• Speech service placeholders are present and will be implemented in a future release.

License

Select a license (e.g., MIT) and update this section.