Skip to main content
New: hear the reviewers' questions before you submit

Submit your paper in a few clicks, to any journal.

Velto keeps every study, file and co-author in one place, checks your manuscript the way the journal's reviewers will, and builds every file and answer the submission site asks for.

Free during beta · No patient data, ever · Every rule linked to its source

Pre-submission review · JAMA, Original Investigation
Fix before submitting 16Head off 5

Methods and statistics

Staged bilateral knees analyzed as independent observationsMethods, paragraph 2Analysis
Complete-case analysis without reporting the extent of missing dataMethods, paragraph 4Analysis
Readmissions to other hospitals not capturedMethods, paragraph 3

Interpretation and claims

Discharge to SNF is likely a marker of frailty, not a modifiable targetDiscussion, paragraph 1
Methods and statisticsLikely revision requestFix: re-run an analysis

Staged bilateral knees and surgeon-level clustering analyzed as independent observations

Two procedures in the same patient share comorbidity and frailty, so treating them as independent makes the standard errors too small and the 95% CIs in Table 2 too narrow.

The change · Refit the Table 2 model as a GEE clustered on patient, add a first-knee-only sensitivity analysis, and report both with 95% CIs.
m <- geeglm(readmit30 ~ age + sex + bmi + diabetes + asa3plus,
            id = patient_id, family = binomial, corstr = "exchangeable")
Velto's review of its sample study (made-up data).

Requirements and submission guides for

JAMANEJMThe LancetThe BMJNature MedicineJBJSCORRJ ArthroplastyBone & Joint J+ any journal

How it works

From draft to ready-to-submit in three steps

  1. Start from a manuscript
    • Title and running title
    • Authors and affiliations6 authors
    • IRB approval and consent waiver
    • Funding and grant numbers
    • 38 references checked2 not found

    Step 1

    Upload your draft

    Velto reads the title, authors and affiliations, abstract, statements and references, checks every reference, and sets up the study. You confirm each field.

  2. JAMA · Original Investigation
    • Main text word limit3,000
    • Abstract350, structured
    • Key PointsMissing
    • Access to data statementMissing
    • Display items5 tables or figures

    Step 2

    Pick a journal

    Its word limits, required statements, upload slots and submission questions, each linked to the journal's own guidelines.

  3. Ready to submit to JAMA
    • 01_Cover_Letter.docx
    • 02_Title_Page.docx
    • 03_Manuscript_Blinded.docx
    • 04_Table_1.docx
    • 05_Figure_1.tiff
    • Only you: sign your disclosure

    Step 3

    Get a ready-to-submit packet

    Every file split, named and formatted the way that journal wants, with the answers to its questionnaire. What's left is what only you can do.

Pre-submission review

Hear the reviewers' questions before they ask them

Five AI reviewers, an editor, a statistician, a clinical reviewer, a reporting checker and a consistency checker, read your manuscript the way the journal you picked would, using what it has recently published. It's a simulation to help you prepare, not a prediction.

  • The questions you didn't see coming

    Clustered and repeated observations, confounding by indication, missing data, outcomes you can't fully capture: the methods points reviewers raise and authors miss.

  • Written like a reviewer, with the actual change

    Each issue says what to change and where to report it, with code to run in R, Stata, SAS, SPSS or Python. Your data never leaves your computer.

  • Ranked for the journal you picked

    What to fix before submitting, what to head off cheaply, and what this journal's recent papers don't usually do.

Methods and statistics reviewer
Methods and statisticsLikely to be raised · Likely revision request · Fix: re-run an analysis

Staged bilateral knees and surgeon-level clustering analyzed as independent observations

Both knees of patients with staged bilateral procedures were analysed as separate cases, but the model doesn't account for the correlation. Two procedures in the same patient share comorbidity and frailty, so the standard errors are too small and the 95% CIs in Table 2 too narrow.

The change

Refit the Table 2 model as a GEE clustered on patient, add a first-knee-only sensitivity analysis, and report both sets of adjusted ORs with 95% CIs.

Run it yourself (R)

library(geepack)
# one row per procedure; cluster on patient
m <- geeglm(readmit30 ~ age + sex + bmi + diabetes +
              asa3plus + discharge_snf,
            id = patient_id, family = binomial,
            corstr = "exchangeable", data = df)
exp(cbind(OR = coef(m), confint.default(m)))

# sensitivity: first procedure per patient
first <- df[!duplicated(df$patient_id), ]

After a rejection

Rejected? Ready for the next journal the same day.

Every earlier submission, decision and reviewer comment stays in the history.

  1. 1

    Forward the decision letter

    Velto files the decision and reviewer comments with the submission, so nothing is lost.

  2. 2

    Pick the next journal

    It suggests journals that fit your paper and your reviewers' concerns, with their limits side by side.

  3. 3

    Get the new packet

    Every file and answer is rebuilt for the new journal, with a summary of exactly what changed.

JAMA → The Journal of Arthroplasty

What changed

  • Abstract: headings changed to the journal's set
  • Highlights: 3 to 5 bullets drafted for you to edit
  • Key Points: removed (not used here)
  • References: restyled for the journal
  • Blinded manuscript: now required

For PIs and labs

Every study in your lab, and who's blocking what

  • Every study in your lab on one dashboard, with a status that updates itself.
  • Who's holding things up: unsigned disclosures, drafts waiting for approval.
  • IRB renewals and training expiry dates, with reminders before they lapse.
  • An authorship plan every author confirms, before it turns into a dispute.
Dashboard
  • Readmission after total knee arthroplasty

    Under review at JBJS · day 23

  • Tranexamic acid in revision hip arthroplasty

    Waiting on 2 authors to approve version 4

    Approve version 4
  • Opioid prescribing after ACL reconstruction

    IRB renewal due in 21 days

    IRB renewal

Trust

Built for research that has to be right

  • Sourced journal rules

    Every requirement links to the journal's official author guidelines, so you can check it at the source.

  • No patient data, ever

    Datasets stay in your institution's approved systems. Velto keeps a link, never the data.

  • Authors sign for themselves

    Each author reviews and signs their own disclosure. Nothing is signed on anyone's behalf.

  • Verified references

    Every reference is checked against Crossref and PubMed, and retracted papers are flagged.

  • AI that's labelled

    Everything the AI drafts is marked for you to review. It never writes your science or invents data.

  • Not used to train AI

    Your manuscript is private to your team and is never used to train AI models.

Pricing

Free during beta

Everything, free while Velto is in beta. No card needed. You'll hear from us well before anything changes.

FAQ

Questions people ask

Is my manuscript private?

Yes. Only people on your study can see it. It's stored privately and sent to the AI provider only to run the checks you ask for, under terms that don't allow it to be used for training.

Does Velto submit for me?

No. Velto prepares every file and answer; you upload them and click Submit on the journal's site, signed in as yourself. Velto never asks for or stores your journal passwords.

Which journals?

JAMA, NEJM, The Lancet, The BMJ, Nature, Nature Medicine, JBJS, The Journal of Arthroplasty, CORR and The Bone & Joint Journal are in the library, with every rule linked to its source. You can add any other journal by pasting the link to its author guidelines.

What does the AI do, and not do?

It reads your draft, checks it against the journal's rules, reviews it the way that journal's reviewers would, and drafts admin text such as cover letters. It doesn't write your science, invent data or citations, or change your manuscript without your approval.

Can I use it with patient data?

No. Velto never holds patient data. Keep datasets in your institution's approved systems; Velto stores only a link and the aggregate results you choose to paste.

Your next submission, without the days of formatting.