Feature

Document analysis

Drop in PDF, DOCX and XLSX together. Brome reads the contract, the invoices and the spreadsheet as one case, then answers from all three at once.

How it works

  1. 1

    Drop the files in

    PDF, DOCX and XLSX, along with images and plain text, in the same conversation.

  2. 2

    Ask across all of them

    Questions can span every file at once instead of one document at a time.

  3. 3

    Trace the answer

    Findings come back attributed to the file they came from.

  4. 4

    Keep working

    Follow up, switch model for a second reading, and the files stay in context.

What it gives you

Several formats at once

A contract, an invoice batch and a forecast read together rather than separately.

Answers with a source

Each finding points back to the document it came from, so you can verify it.

Second opinions

Hand the same documents to another model without uploading them again.

Not training data

Zero data retention with every provider, hosted in the EU, never used to train models.

Put the whole file set in

Stop summarising documents one at a time to something that can only read one.