Skip to main content

AI service

Document visual question answering

We provide Document visual question answering at any scale, on a private cluster dedicated to you — from a single GPU to a hundred, on premise or in any cloud.

If you need such a service please contact us.

Ask a plain question of a page and the answer comes back with the region it was read from, so it can be checked rather than trusted. It works on the page image itself, which means charts, stamps, handwritten notes and table cells are all in scope. Multi-page documents are handled as a unit, so an answer can be drawn from a table on one page and a note on another.

Capacity can be permanent or temporary: a one-off archive can run on burst capacity and release it afterwards, while a continuous workload sits on a cluster reserved for you. Our AI advisors help pick the model and the hardware before anything is committed.

More services