Conference peer review software

Double blind

Reviewers see the title, abstract, track and keywords — never the authors’ names or affiliations. The names are withheld before the reviewer’s page receives the submission, not hidden by the screen. This is the default.

Single blind

Reviewers see who wrote the submission. Their own identity is never shown to authors.

Open review

Reviewers see the authors, for committees that review in the open.

Not sure which fits your meeting? Single-blind vs double-blind review compares the three.

You add a reviewer with a name, an email, an institution and up to 20 topics. They receive an invitation by email for this event only. Signed in, a reviewer sees one page: their assignments, the deadline, and a scoring form for each. They cannot see the rest of your workspace, other reviewers’ scores or submissions assigned to someone else.

If a reviewer recognizes a conflict you could not have known about, they declare it and the assignment is released, so you can give it to someone else.

Set how many reviewers each abstract needs — one to ten, three by default — and a review deadline. Then assign in whichever way suits your committee:

Manual

Pick reviewers for each submission from the queue.

Automatic

A fixed rule fills the gaps: reviewers whose topics match the submission’s track and keywords first, then whoever has the lightest load. It is a rule, and it is labelled as one.

AI-assisted recommendations

AI suggests reviewers for submissions that still need them, with a short reason for each match, choosing only among reviewers the conflict rules allow. It writes nothing: you confirm the suggestions in one step, or change them first. The AI is given each submission’s title, track and keywords and each reviewer’s topics and workload — never names, email addresses, institutions or author lists.

• A reviewer is never assigned a submission on which they are the submitter or an author.
• With “skip same institution” on, a reviewer is not assigned work from an author at their own institution. Institutions are compared after ignoring capitals and punctuation, so “Heart Institute” and “heart-institute.” match — but two different universities are never treated as one.
• With “reviewer declares” on, reviewers can decline an assignment by declaring a conflict.

These rules apply to every route into an assignment — by hand, by rule and to AI suggestions — so a conflicted pairing cannot be created by choosing a different button.

Each review scores every criterion from 1 to 5. The defaults are scientific quality, originality, relevance and methodology; you can rename them or use up to ten of your own. A review also carries a recommendation — accept, request revision or reject — and two comment boxes: one for the authors and one for the committee only.

The conference review scoring rubric describes what each score means for each default criterion, so reviewers calibrate before they start.

The decisions queue shows each submission with its reviews, the average score and each reviewer’s recommendation. An AI summary of the reviews can be generated for each one, and underneath every summary the screen says the same thing: “AI summarizes the reviews. You make the decision.” No step of the product accepts or rejects anything on its own.

A committee member chooses Accept, Reject or Request revision, with an optional note. The decision is recorded with who made it and when. If a review changes after a summary was generated, the stale summary is not shown.

When decisions are final, you publish the accepted abstracts. Publishing is what lets them appear in the abstract book and on your website — and a published abstract’s decision can no longer change, so what the public has seen stays true.

An accepted submission becomes a presentation — oral or poster, following the author’s preference unless you change it — ready for papers and posters to be collected and for the program to place it. Running review for the first time? How to organize peer review for a conference walks through the whole process.