How can we help?

Choose Register and fill in your details, or continue with Google. An account is free. You need one to publish, react, comment, collaborate or use Check Idea; anyone can read published research without one.

So that every author, comment and collaboration request on DataIntell comes from a real address. After you register we email you a link; open it on any device and your address is confirmed. Accounts created with Google are confirmed already.

Until then you can read, browse and write drafts, but not submit research, comment, request collaborations, invite co-authors or apply to review. If the link has expired or did not arrive, open Confirm your email and send a new one, and check your spam folder.

Your name, a short bio, your research interests and your institution. Your profile is public: it lists your published research, co-authored work and impact. Pick your institution from the list when it is there, so that reviewers from your own institution are kept off your work.

Changing your email address means confirming the new one, so we can still reach you.

Your workspace brings together your projects and drafts, your library of papers, the Ongoing Research Registry, your extensions and your collaboration requests. The bell in the navigation shows new notifications.

Choose Publish research. The wizard has four short stages and a review page:

1. Problem: title, field and the problem your research addresses.
2. 4US scoring: how worthy the problem is, with a justification and sources.
3. Methodology: how you did the work, and any paper or files.
4. Results: what you found, its limitations and any product built from it.

Your draft saves as you type, so you can stop and come back. On the review page you check everything, add co-authors and choose how your work is validated. It goes live as soon as you submit.

Enough for a reader and a reviewer to understand the work. The minimums are a title of 10 characters, a problem description of 100, a methodology and results of 150 each, and limitations of 50. Each 4US justification needs 50 characters, or 100 when you score that dimension 4 or 5. A counter under each box shows how far you are.

Four questions that test whether a problem is worth solving, each scored from 0 to 5:

• Unworkable: do current solutions fail?
• Unavoidable: does the problem affect many people?
• Urgent: does it cause harm that is time-sensitive?
• Underserved: are there gaps in what exists today?

The total is out of 20, and 10 or more meets the threshold for a worthy problem. Score honestly and back high scores with sources: reviewers check the scoring, and a well-justified 12 reads better than an unsupported 18.

On the review page, open Co-authors and add each person's name, with their email and institution if you have them. Everyone with an email is invited when you submit. They are credited on the project and on their profile once they confirm, and anyone not involved can ignore the invitation. You can invite people without an email later, from the project page.

Yes. From Import you can bring in a paper by its DOI, upload a PDF, or pick from your ORCID record. DI_AI, our research assistant, maps the paper into a draft for you to check and complete before you submit. Co-authors found in the paper are listed for you to invite when you are ready.

It helps. Check Idea looks for published projects, registered ongoing research and wider literature close to your idea, and tells you how close the nearest work is. Similar work never stops you publishing: a different approach, place or population is a real contribution. It helps you explain what is new, or find someone to work with.

You can update the methodology, results, files and product details at any time. The title and problem description are fixed once you submit, because your proof of authorship and the reviews are based on that wording. To change them, start a new version from My projects: it links back to the original, which keeps its certificate and history. If validation finds your submission needs work, everything stays editable until you resubmit.

Yes, from the project's page or from My projects. Deleting removes it from DataIntell for good, including its certificate, so download anything you need first.

When you submit you choose AI only, Human only, or AI and human, which we recommend. AI validation gives detailed feedback quickly. Human review adds the judgement of approved reviewers and the stronger badge. Your work is public and readable throughout, whatever you choose.

AI validation usually finishes within minutes. Human review: 2 approved reviewers review your work in turn, and each has 72 hours once they take it on. If they disagree, a third reviewer decides. How soon they start depends on how many reviewers are available in your field.

We email you when a reviewer starts, when each review is in, and when the decision is made.

Approved reviewers take projects from a queue, and your work is shown first to reviewers who cover your field. Each reviewer scores it against a 100-point rubric (problem, methodology, results and impact) and recommends an outcome. 2 reviews are needed.

Reviews are blind: reviewers see your work without your name, and you see their feedback without theirs. Reviewers never review work by their own institution or work they helped write. If the reviews disagree, a third reviewer decides; if there is still no majority, the DataIntell team reads every review and decides.

• AI validated: passed AI validation.
• Human validated: approved by human reviewers.
• Human + AI validated: passed both, the strongest validation badge.
• Externally reviewed: already peer-reviewed by a recognised publisher.

A project without a badge is still published and readable; it simply has not been validated yet.

The highest recognition on DataIntell, awarded only by human reviewers: most of a project's reviewers must recommend it for Showcase, with an average of 85 or more out of 100. Showcase projects get a Showcase certificate. Your own 4US score never makes a project Showcase.

You get feedback that explains what to improve. Your project stays published while you work on it, and you can edit it and resubmit. DataIntell is not a rejection system: the aim is better research.

Read the feedback carefully, then contact us with the project and what you think was missed. The team looks at every request.

Every published project gets a cryptographic fingerprint of your authorship: your name, the title and the time you published. It is also anchored to the Bitcoin blockchain, usually within a day, which makes the date independently provable. It is lasting evidence that you published this work first.

• Proof of Authorship for every published project.
• Validated Research once your project passes validation.
• Showcase when reviewers award it.

Find them on your project page. Each has a QR code and link to a public verification page, ready for LinkedIn, applications and portfolios.

Through the verification link or QR code on the certificate. The page shows the project, author, date and fingerprint. The timestamp proof can be downloaded and checked on opentimestamps.org, without trusting DataIntell at all.

No. It shows that your work was published, and validated where it says so, on DataIntell. It is not peer-reviewed journal publication and should not be presented as one.

Points come from what people do with your research:

• Readers: views count on a diminishing scale, so 100 views earn 20 points
• Each reaction: 10 points; each comment: 15
• Each active collaboration: 30; each completed one: 75
• Each time another researcher's project cites or extends yours: 50

Validation adds up to 30 points, an external peer-reviewed publication up to 40, and a live product built from the research can add several hundred. On your own project page, the Research impact box shows exactly how your score is made.

Views count real readers. Your own visits, your co-authors' visits, repeat visits by the same person on the same day, and automated traffic such as search engine crawlers and link previews are not counted, and we review counts from time to time to remove anything that is not a reader. Paying for traffic or inflating numbers is against our terms and never helps: it is removed.

The leaderboard ranks projects and researchers by impact, with boards for this week, rising work, the most discussed and the most reacted, and the top project in each field. The best way up is research people engage with.

Authors can mark a project as seeking collaborators and say which skills they need. On a project page, choose Request to collaborate and write a short proposal. If the author accepts, a private conversation opens. When the work is done the author marks it complete, which counts towards the project's impact.

You register that you are building on someone's published work: replicating it, applying it somewhere new, or taking it further. The original author is told and credited. When your work is published, link it to the extension, and the two projects show their connection. No permission is needed: building on research is open by design. You cannot extend your own project; start a new version instead.

A way to stake your claim to research before it is finished. Register your topic and problem, and you get a timestamped proof that you were working on it from that date. Public registrations appear in discovery and in Check Idea, which helps you find collaborators; an embargoed registration is visible only to you until you publish. You can have two active registrations at a time, and we remind you as your expected completion date approaches.

Be constructive and specific, and discuss the research, not the person. You can edit or delete your own comments. Harassment, spam and abuse break our terms; please report anything that does.

Apply from Become a reviewer. The page shows which criteria you already meet. There are two tiers:

• Community Reviewer: at least 3 published projects, 2 of them human validated, an average 4US score of 10 or more, and an institution on your profile.
• Expert Reviewer: either 5 published projects with 3 externally peer-reviewed, or 8 published projects with 4 human validated and an average 4US score of 15 or more; plus an institution.

The team reads every application and emails you the decision.

Open your reviewer dashboard and confirm your institution first: it keeps you off projects from your own institution. Then take a project from the queue, declaring that you have no conflict of interest.

72 hours from taking a project on. If the time runs out, the project goes back to the queue for another reviewer, so only take on what you can finish.

Score each part of the rubric on the evidence in the submission, name real strengths, and make every weakness specific enough to act on. Recommend Showcase only for exceptional work. Remember the author will read every word.

Use Forgot your password? on the login page. The reset link works once and expires after a few days. If you usually sign in with Google, keep using Google.

For your security, you are signed out after a period without activity. A countdown appears first, with a button to stay signed in, and drafts you are writing are saved before you are signed out. Log in again and you return to the page you were on.

Emails about your account and your own work always reach you: confirming your email, password resets, publication, and each step of your validation and review. Other emails are yours to choose: comments, collaboration, research that builds on yours, milestones, people you follow and registry reminders. Turn any of them off in Email preferences, or with the link at the bottom of any of those emails. Everything still appears in your notifications.

Published research and your public profile are visible to everyone. Your email address, drafts, collaboration conversations and embargoed registrations are private. See our privacy policy for the details.

Email privacy@dataintell.co.uk from the address on your account. We remove your personal data within 30 days.

Reset your password straight away, then contact us so we can look into it.