API Schema Extractor: Automate API Discovery & Schema Generation

This sophisticated workflow automates the end-to-end process of discovering APIs from various services and generating custom schemas in three stages: research, extraction, and schema generation. In the research stage, it fetches pending services from a Google Sheet, uses Google search to locate API documentation, employs Apify for targeted web scraping, stores webpage contents and metadata in a Qdrant vector database, and updates progress status (pending, ok, or error) in the sheet.

The extraction stage processes successfully researched services by querying the vector store for products, offerings, and API docs, then leverages Gemini LLM to extract API operations, recording them in the Google Sheet while updating status. Finally, the generation stage retrieves all operations, combines and groups them into a custom schema, uploads it to Google Drive, and logs the file location in the sheet.

Ideal for development teams cataloging multiple APIs, API documentation initiatives, or standardizing schema collections. It saves significant time on manual research and extraction, enabling scalable automation with high accuracy via AI and vector search. Requires Google, Apify, Qdrant, and Gemini accounts.