Peer Review

Peer review, with Alfred on the reviewer list.

Assign your manuscript to the people who will read it, and add Alfred to the same list. It checks the citations it can reach, reads your claims against the sources behind them, and questions the argument, so the human review can go on the things only a human can catch.

Start 3 days for ₹48Try Pro free for 7 days Explore all features

3 days of Pro for ₹48 · then ₹899/mo · cancel any time7 days of Pro, no card · then the free plan · upgrade when you want to

How it works

Three steps from a finished draft to a review

The same flow whether the reviewer is a person or Alfred.

1

Mark the manuscript Final

Finish the draft and mark it Final. The manuscript stops moving under the reviewers, so everyone is reading the same version and every comment stays anchored where it was left.
2

Assign your reviewers

Add the people you want: your guide, a colleague, a lab member with the Reviewer role. Then add Alfred to the same list. Alfred does not take one of your paid reviewer seats.
3

Triage what comes back

Human comments arrive as threads. The findings from Alfred arrive in a triage panel, grouped by category and severity. Accept the ones worth keeping and each becomes a comment thread on the passage; dismiss the rest and the manuscript is untouched.

What Alfred reads for

Four things a reviewer would check first

Citation integrity

Markers pointing at a source that no longer exists, the same paper cited twice under two records, DOIs that do not resolve, a manuscript that cites sources but carries no reference list, and citation styles that drift between two forms. For a reference list you typed by hand, it also finds entries you never cited and in-text markers with no matching entry.

Claims against the sources they cite

For a claim that cites a source, Alfred reads the source and says one of three things: the source supports the claim, the claim goes further than the source does, or the source does not support it. This is the check that catches a citation quietly stretched over two revisions.

Argument and method

Read one section at a time: claims asserted without support, conclusions the evidence does not reach, contradictions between sections, design choices left unjustified, limitations that are missing, assumptions never stated.

Writing

The same checks the manuscript check runs. Spelling that drifts between two forms, jargon with a plain equivalent, genuinely wordy phrasing, phrasing that reads as AI, typography, and abbreviations used before they are defined.

There is no score and no verdict at the end of it. A review is something you read and judge, not a number to clear.

What is not checked

What Alfred cannot check, said plainly

Alfred can only check a citation when it can actually reach the source. That means one of three things: the paper is in your own library, or it is open access, or its abstract is reachable. When none of those work, Alfred reports the citation as unchecked in the coverage panel and produces no finding about it.

So a review never implies Alfred read a paper it could not reach, and it never claims to have checked every citation. The coverage panel is where you find out which ones it skipped, and the honest answer for a paywalled journal article you have no copy of is that Alfred has not read it.

This is the same discipline the originality check follows: it searches open-access literature and the open web, not paywalled journal archives, and says so. A review you can trust is one that tells you where it stopped looking.

None of this replaces your supervisor or the journal's reviewers. It is the pass you run before you take the draft to them, so their time goes on the things only they can catch.

The whole workflow

The review your readers actually run

Reviewers, not co-authors

Invite a reviewer to your workspace with the Reviewer role and they see the manuscripts you assign them, in a read-only view, and nothing else in your library. Your guide, an external examiner, or a lab member can read and comment without being given the run of your research.

Inline comments on the passage

A reviewer selects a passage, picks a highlight colour, and writes. The highlight marks the text so you can see at a glance where the reading slowed down. Comments that are about the whole manuscript go in the general comments panel instead.

Threads, so a point can be argued

Every comment is a thread. You reply, the reviewer replies back, and the discussion about one sentence stays attached to that sentence rather than scattering across email.

Resolve as you revise

Address a point and resolve the thread. Resolved comments are grouped away and dimmed, so what is left in the panel is what is left to do. You can unresolve one if the revision did not hold.

Numbered rounds

Each Alfred review is a numbered round, stamped with the manuscript version it read. Round two sits beside round one, so the record of what you fixed is the review itself rather than your memory of it.

Export the review

Export or print the manuscript from the review page as a PDF, for your records, your committee, or the file you keep for the submission.

"The mechanical faults were gone before my guide opened the file."

What is included

What each plan includes

Pro

Up to 3 reviewers, and 3 Alfred review rounds a month. Alfred does not use a paid reviewer seat.

Lab

Unlimited reviewers, and 10 Alfred review rounds a month, across the whole group.

Alfred as a reviewer is a paid feature. It is not available without a paid plan.

Where to next

Before you submit

Go deeper

Review the draft before anyone else has to.

Mark a manuscript Final and add Alfred to the reviewer list.

Start 3 days for ₹48Try Pro free for 7 days See pricing

3 days of Pro for ₹487 days of Pro, no card Then ₹899 a monthThen the free plan, upgrade when you want to Cancel any time

Prefer to read first? Read the Docs