Unstructured builds tools that convert documents such as PDFs, HTML pages and Word files into structured data for language models and retrieval pipelines. Its core is an open-source library released under the Apache 2.0 licence. The company also offers a hosted platform and API for higher-quality processing, and connectors to enterprise sources. Deployment can be cloud or private. Unstructured Technologies, Inc. is headquartered in the United States.
As a US company, Unstructured can be subject to requests from US authorities for data under its control. Its privacy notice says it may use service providers in the United States and other countries, without naming EU hosting locations. Transfers rely on European Commission adequacy decisions and the transfer mechanisms of GDPR Article 46, and the company states that it is GDPR compliant. An organisation can reduce its exposure by running the open-source library on its own infrastructure, so processing stays there. For the hosted platform, the company's processing terms apply, which it says govern content processed on behalf of customers.