Skip to main content
Connect your Azure OpenAI or Azure AI Foundry agent to Doconda so it can convert, create and edit Word, PowerPoint, Excel and PDF files. The SDK tools plug into function calling with AzureOpenAI in Python or @ai-sdk/azure in TypeScript.

Capabilities

  • Create a Word, PowerPoint, Excel or PDF file, or an artifact, from a request or from Markdown the agent wrote.
  • Edit a file: only what is asked changes, and get_report lists every change.
  • Convert a file to another format with no AI, for example any file to Markdown so the agent can read it.
  • Read what people saved in an artifact, such as poll votes.

Setup

This integration uses a Doconda API key in DOCONDA_API_KEY (from API keys). Azure credentials come from your environment, as in the rest of your code. model is the name of your Azure deployment.

With Python

AzureOpenAI() reads AZURE_OPENAI_ENDPOINT, AZURE_OPENAI_API_KEY and OPENAI_API_VERSION. The last message includes the report’s download links.

With the AI SDK

azure reads AZURE_RESOURCE_NAME and AZURE_API_KEY.

Azure AI Foundry

Any Foundry deployment that supports function calling works the same way: the tools are standard JSON Schema.

Tools

The agent gets the eight Doconda tools; see the tools for what each one does. doconda.tools() returns them with name, description, a JSON Schema (inputSchema in TypeScript, input_schema in Python) and run. A failed call returns { "error": "<code>", "message": "…" } instead of raising, so the model can correct itself.

Troubleshooting

  • Doconda: pass api_key or set DOCONDA_API_KEY. — The SDK found no key (in TypeScript: Doconda: pass { apiKey } or set DOCONDA_API_KEY.). Set DOCONDA_API_KEY, or pass the key to Doconda.
  • The model never calls a tool — Check that the deployment’s model supports function calling.
  • A tool result has "error": "insufficient_balance" — The project’s balance can’t cover the document: top up in the dashboard, or ask for quality: "fast".
  • The download link stopped working — File links expire after 5 minutes; the agent calls get_document for fresh ones.