sifter-ai/sifter

What it does

Sifter is an open-source document extraction and structuring tool that converts unstructured documents into typed, schema-defined records. It enables users to query documents as if they were databases using natural language, making complex document analysis accessible through conversational interfaces.

How to set up

Sifter operates as an MCP (Model Context Protocol) server that integrates with Claude, ChatGPT, Gemini, and other MCP-compatible clients. Installation involves cloning the repository from GitHub, installing Python dependencies, and configuring it as an MCP server within your preferred AI platform. Detailed setup instructions are available in the project's GitHub repository.

Key features

• Converts unstructured documents into typed, structured records
• Schema-defined extraction for consistent data output
• Natural language querying of document contents
• Compatible with multiple AI platforms and MCP clients
• Open-source implementation for transparency and customization
• Database-like query capabilities for processed documents