Terms of use.

What the service is

Prizmatic is an evidence platform. It searches the literature databases you choose, screens and classifies the records it retrieves, extracts data from the studies that pass, and produces a study table, an evidence grid, a written synthesis and a PRISMA-style record of every decision. Related tools read a finished review to support a grant application or to check a set of specific aims.

What the service is not

It is not an author, a reviewer or a guarantor. As the FAQ says, you remain the author and the guarantor of any review you register or publish, which means checking what the pipeline produced rather than forwarding it. The audit trail exists so that the check is possible; it does not replace it. Nothing on this site is medical advice, and no output should inform a clinical decision without a qualified person reading the underlying studies.

Accounts

Some parts of the service require a signed-in account. Sign-in is handled by Clerk, and you are responsible for what is done under your account and for keeping your credentials to yourself. Each review belongs to the account that launched it, and the backend refuses to show it to another account. Some tools are open only to particular accounts, and we may change which. We may suspend an account that breaks these terms.

Acceptable use

• Use the service for research and evidence work, and to judge whether it fits your needs.
• Do not try to reach another person's reviews, guess job identifiers, or get around sign-in.
• Do not point the service at content you are not permitted to process.
• Do not use the literature services through us in a way that breaks their own terms, such as bulk harvesting.
• Do not probe, overload or attack the service. There are rate limits, and we may block traffic that trips them.

The demos

The public demos replay a review that was captured once. The query builder may run live, but the pipeline, the table and the synthesis are frozen output from that capture, not a fresh run. They show what the service produces; figures in a demo may not match a new run of the same question.

Your content and our outputs

Your research question, your settings and anything else you enter remain yours. You give us permission to process them, and to send them to the providers named on the privacy page, for the purpose of producing your review. The outputs produced for your review are yours to use; we make no claim on them. Retrieved records remain subject to the terms of the database they came from.

We may look at runs to operate, debug and improve the service. Our own staff can see a review's title, who launched it and its files through an internal console.

No warranty of accuracy

The service is provided as is. Models make mistakes, and we do not yet publish a measured rate for how often. We do not warrant that a search is complete, that a screening decision is right, that an extracted figure is correct, or that the service will be available at any given time. Checking the output is part of using it.

Limitation of liability

To the extent the law allows, we are not liable for indirect or consequential loss, or for any loss arising from reliance on an output that was not checked. Where liability cannot be excluded, it is limited to the amount, if any, you paid for the service in the twelve months before the claim.

Changes to these terms

We will change these terms as the service changes. When they take effect the date will be at the top of this page, and a material change will be announced on the site before it applies. Continued use after that date is acceptance of the new terms.

Contact

Contact: to be added.Open item