Unsiloed Inc. has secured $7.5 million in a seed funding round led by Felicis. Other participants included Y Combinator, Orange Collective, Uncorrelated Ventures, Aarthi Ramamurthy, and Beth Turner.

The company provides a platform that converts complex documents, such as PDFs, financial statements, and clinical records, into structured data formats like Markdown or JSON. This technology is designed to make unstructured information readable for AI agents.

The platform utilizes a dual-stream vision architecture that processes document content and visual layouts simultaneously. This approach preserves relationships between headings, charts, and text sections, which standard OCR technology often fails to capture.

Unsiloed reported a score of 88.0 on the olmOCR-Bench benchmark. This result places the system ahead of other models, including OpenAI's GPT 5.5 at 84.6, Anthropic's Claude Opus 4.7 at 81.9, and Google's Gemini 3 Pro at 77.7.

The company has processed millions of pages for various clients, including AI startups, banks, insurance firms, and law firms. The platform supports deployment on private clouds, on-premises, and air-gapped infrastructure.