What ByteRover is
Search that shows the passage and the page. Ask in plain language, and get one cited answer, or "not found" when your files do not answer.
ByteRover searches the files your team puts in a drive. Ask a question in plain language and you get back the passages that answer it, each with its page and heading, so you can check every one against the file.
Choose your start
Ask in the app
Upload your files and ask your first question, without writing code.
Connect your agent
Give Claude, ChatGPT, Cursor or any other agent access to your team's files with one address.
Build with the API
Upload a file and search it in two calls.
All three search the same drives and return the same passages.
What you get back
You ask: Which policy section covers EU refunds?
The EU refund rules are in the customer refund policy, under Refunds > European Union. Customers have 14 days to cancel. [1]
[1] customer-refund-policy.pdf · page 6 · Refunds > European Union
"Customers in the European Union may cancel an order within 14 days of delivery and receive a full refund. The refund goes back to the original payment method within 10 working days."
| You get | What it is |
|---|---|
| Passages | Up to 10 passages, best first. Each has its text, its page, slide or sheet, and the headings above it. |
| A cited answer | A short answer written only from those passages, in the app. Each sentence that states a fact names the passages it uses. |
| "Not found" | A clear result when your files don't answer the question. ByteRover doesn't guess. |
If you ask Is there a policy on drone inspections? and no file covers it, the result is "not found". That tells you the gap is in your files. Results covers all three in detail.
How your files are organized
Team Acme Robotics members, API keys, billing
└─ Drive Policies who can read files and add files
└─ Folder /policies/eu a path that can limit a search
└─ File customer-refund-policy.pdf revisions, tags, statusPeople see only the drives they've been given, and a search only runs on drives its caller can use. See Teams, drives and folders.
Good to know
There's no embedding pipeline to run. A file is ready to search once its status is Searchable, and when it changes you upload it again. There's nothing to re-embed.
Your files are not training data. Nothing you upload is used to train a model. See Security and data.
ByteRover reads PDF, Word, PowerPoint, Excel, HTML, Markdown, text, email and code files, up to 1 GB each. Very large Word, PowerPoint and Excel files can hit a reading limit first. ByteRover can't read a PDF that's only images yet, but support is on the way. See Files.