Tennr

Tennr builds automation for the processing of patient referrals in American healthcare. Its systems use a vision-language model to read, extract and act on referral documents, covering the workflow from intake through to prior authorisation. The problem it addresses is substantial: by the company's cited industry figures, some 90% of provider-to-provider communication in the United States was still conducted by fax as of 2024, and referral packets commonly arrive as 20-plus-page PDFs containing handwritten notes in thousands of different formats.

The core of the technology is RaeLM™, a vision-language model trained on more than 3m medical documents. It classifies incoming documents, extracts the relevant clinical and administrative information, populates electronic health records without human intervention, verifies benefits and manages the associated back-and-forth correspondence. Tennr processes more than 10m documents a month through this pipeline.

The technical work spans vision-language modelling, document understanding and OCR, intelligent document processing, EHR integration and workflow automation. The company serves provider networks, medical practices and specialist referral operations, as well as revenue-cycle and prior-authorisation teams, across the United States.