Free to start · Scanned PDFs on Pro

Chat with any PDF.
Verify every answer.Ask questions about the PDF you have open, from contracts and papers to scans and files behind a login. Every answer links to its sentence.

Aurora Rollout Runbook

Overview

Releases leave the main branch every Wednesday at 14:00 UTC. A release candidate is cut automatically, tagged RC, and deployed to the staging cluster within fifteen minutes. Nothing reaches production without passing the full staging suite twice in a row.

Approvals

Every production rollout needs two approvals: the release captain on duty and one engineer from the owning team. The release captain checks the error budget before signing off; if more than 25 percent of the monthly error budget has been burned, the rollout is deferred to the next window.

Canary stage

The canary stage routes 5 percent of production traffic to the new build for 45 minutes. The canary cohort is selected by session hash, so a returning user stays on the same build for the whole window.

Rollback

Rollback is one command: aurora rollback --to-last-good. It repoints the fleet at the previous build and completes in under four minutes. Run it from the deploy bastion, never from a laptop.

Answer

This runbook describes how the Meridian web app ships to production through the Aurora pipeline: a weekly release cut , a two-approval gate , an automated canary stage , and a single-command rollback path .

How do I chat with a PDF and get answers I can check?

Open the PDF in your browser, open the Sidenote side panel and ask your question. The answer cites the exact passage it came from: click a citation and the PDF jumps to that page and highlights the sentence. It works on textbook chapters, contracts, research papers and scans, and on PDFs behind a login, with no upload step. Studying? How to chat with a PDF textbook.

Live demo

Try it on a contract. No account needed.

Paste a link or upload a small file. The engine behind the extension returns a cited summary: click a chip to read the passage it came from.

Why it's different

A PDF answer you can't verify is just vibes.

Upload-first PDF chat tools hand you plausible prose you can't check. Sidenote validates every claim against the document on the server, drops what it can't back up, and lets you click through to the evidence.

Answers with receipts

Every answer carries numbered citations. Click one and the PDF scrolls to the exact sentence, highlighted - claims that can't be backed by a passage are dropped server-side.

Scanned PDFs on Pro

On paid plans, image-only PDFs - scans, faxes, signed contracts - are OCR'd first, so even paper gets real, citable answers.

Behind-login PDFs just work

A PDF in Drive, SharePoint, your wiki or an email attachment is read in place through your own signed-in session. No download-and-re-upload dance.

Chat across many PDFs

Group PDFs into a collection and ask one question across all of them - the answer cites whichever document each point came from.

Deep-dive: how Sidenote reads PDFs · how citations are verified · chat with documents beyond PDFs

FAQ

Chatting with PDFs, answered.

Yes. Open the chapter PDF in your browser and ask in the Sidenote side panel. Every answer cites the passage it came from, and clicking a citation jumps the PDF to that page and highlights the sentence, so you can quote the page number with confidence. For a whole textbook, group the chapter PDFs into a collection (paid plans) and ask across all of them.
Yes. Sidenote's free plan has no time limit and needs no card. It includes 10 cited chat turns a week plus 2 documents, 6 summaries, 6 explanations and 1 glossary build a week. There's also a free instant summarizer at getsidenote.app/tools/summarize that needs no account at all.
Two ways. First, citations: every answer is checked server-side against passages retrieved from the PDF, unsupported claims are dropped, and clicking a citation scrolls to the exact sentence. Second, no upload step: Sidenote is a browser extension, so it chats with the PDF you already have open - including files behind logins that upload-first tools can't reach.
Yes. Scanned and photographed PDFs have no text layer, which still trips up many upload-first tools. Sidenote runs OCR first (paid plans), turning the scan into readable, citable text you can click through to.
Yes - that's the point of living in your browser. A PDF in SharePoint, Google Drive, Confluence or attached to a Gmail/Outlook message is read through your own signed-in session, read-only. If your account can open it, Sidenote can read it.
Free accounts cover PDFs up to 5 MB; paid plans raise that to 20-75 MB by tier. Long PDFs are chunked and embedded, so a 200-page document is fine - answers retrieve the relevant passages rather than stuffing the whole file into a prompt.
Yes. Add them to a collection and ask across the whole set - each answer's citations name the specific PDF and jump to the passage inside it.
Ready when you are

Stop digging. Start asking.

Add Sidenote to your browser, open any page in your wiki, and ask it the question you’ve been Slacking the team about.

Free plan · No card required