Built in the Gulf, for the Gulf
Peer review your journal can defend.
Submission, review, decision and publication in one system — in Arabic and English, with a record that holds up when somebody asks how a decision was reached.
Both native, both first-class. Right-to-left throughout, and search that copes with unvowelled spellings.
Residency is enforced at the network layer — a call that would send manuscript text abroad is refused, not logged.
Workflow and audit records are append-only and linked. Tampering shows.
One link carries accept, read, review and discuss. Reviewer friction is why invitations get ignored.
Who it is for
The same record, read four different ways.
Publishers and societies
One place to run every title, without a per-journal rebuild.
Journals, sections, collections, boards and issues per title. Each journal keeps its own workflow shape, its own thresholds and its own public site.
Editors
To decide well, quickly, and be able to explain it later.
A desk sorted by what needs you. Reviewer consensus read across the reports. Whether the revision answered each point. A decision letter composed with the reviews attached.
Authors
To submit without a fight, and know where they stand.
Drag the file you already have — no template. A writing check as you type. A status page that says what is happening and what comes next.
Institutions and IT
Answers for a security questionnaire.
SAML and OIDC single sign-on, SCIM provisioning, per-tenant network allowlists, encryption at rest with your own key, legal hold, and a support-access log the customer can read.
The part nobody else has
Reading aids that need no AI at all.
Where the reviewers agreed, whether a response letter answered each point, how a submission compares with what the journal actually publishes — all computed, on every plan, for every journal. Nothing leaves your building and nothing is invented: every line is a sentence a reviewer really wrote, quoted exactly.
- Reviewer consensus
What more than one reviewer raised, and where they pull apart — quoted exactly, never paraphrased.
- Revision check
Each point raised, lined up against the response letter. The unanswered ones listed on their own.
- Writing check
Runs in the author's browser as they type. Over-long sentences quoted, abbreviations never spelled out, structure, length.
- Scope comparison
How a submission sits against what this journal has actually published, with the closest articles named.
A language model is available as an add-on for the small residue that genuinely needs one — and it is off unless you turn it on.
What Colophon does
Everything an editorial office needs — nothing that gets in the way
Candidates ranked by real overlap with the manuscript, with the reason shown. Conflicts screened against the record — co-authorship, shared affiliation, recent collaboration — and stored, so a challenge months later can be answered with what was known at the time.
Every point the reviewers raised, lined up against the response letter, with the passage that answers it quoted. The ones nothing touches are listed separately. It never rules on whether an answer was good enough — that judgement is the editor's, and a fluent non-answer looks identical to any machine.
Every listed co-author is asked whether they consented to be listed. A decline goes to the editor alone, never back to the submitting author. Named authorship on work somebody never saw is how a large share of paper-mill output moves.
The article stays published and stays retrievable — that is required, not a limitation. A reader following a citation finds the paper and the notice. The worst notice wins the banner, not the newest.
Workflow events and audit entries are hash-chained and append-only. When an appeal, an ethics case or an auditor arrives, who did what and when is a fact you can produce rather than an argument you have to make.
Crossref DOI registration, OAI-PMH for harvesters, per-journal sitemaps, and a public article page built for the way people actually arrive — from a citation.
The workflow
From submission to decision, in one place
- 1Submit
Drag the file you already have. Declarations captured up front, screening runs automatically.
- 2Assign
Screening results and a scope comparison in front of the editor before anybody is invited.
- 3Review
Invite by link. Independent reports first, then an interactive forum if the journal wants one.
- 4Decide
Consensus read across the reports, then a letter that releases the reviews meant for the author.
- 5Revise
A point-by-point response, checked against what was asked, and back into review as a new version.
- 6Publish
Production, proof, issue, DOI — and a public page the indexes can read.
Sovereign by design
Your manuscripts stay in region, and every decision stays defensible.
A strict data-residency boundary keeps an unpublished manuscript in region — enforced at the network layer, not promised in a policy. Every state change is written to a hash-chained audit trail, so an ethics investigation or an appeal turns on a record that cannot be quietly rewritten.
Run a journal people take seriously.
Bring one journal or forty. Import what you have, keep your DOIs, and be running next week.