SyRev - AI pre-screening that cuts the pile before Covidence
by•
AI pre-screening: SyRev auto-excludes clear nos against your criteria, then exports remaining records as RIS for Covidence. 96.5% of relevant studies kept across 310 public datasets; weak results published. 2 free title/abstract evaluations.

Replies
Hi Product Hunt 👋
Screening is the slowest part of a systematic review. A single search can return thousands of titles and abstracts, and most are clear nos. Reviewers still have to read every one, and if AI helps, they have to be able to justify it to peer reviewers.
That's why we built SyRev. You give it your search and your eligibility criteria. It scores each record against each criterion and auto-excludes only the records very unlikely to qualify (our default "Careful" setting: an estimated chance of inclusion below 0.1%). Everything uncertain, and every candidate include, stays with your reviewers. You don't label any training records first.
If you already screen in Covidence, SyRev is meant to sit in front of it, not replace it. Many teams pre-screen in SyRev, then export the remaining records as RIS and import them into Covidence for their usual dual screening. SyRev is not affiliated with or endorsed by Covidence.
What makes it different: we publish how well it works, including where it doesn't. At Careful, SyRev kept 96.5% of relevant studies across 310 public review datasets and auto-excluded 74–79% of records in full search pools from published systematic reviews. In 31 of those 310 datasets it kept fewer than 95%. We list those reviews and explain why (usually how the criteria were worded): https://syrev.ai/validation
Every review also gets a validation pack you can report: a blind sample of the AI's exclusions with a confidence interval, a PRISMA 2020 diagram, a PRISMA-S search log and a methods paragraph.
Who it's for: PhD students, academic review teams, librarians and evidence-synthesis services, and HEOR / consulting teams (on commercial pricing).
Pricing: every new account gets 2 free title/abstract evaluations. The best test is a review you've already finished, so you can compare SyRev's decisions with your own and see how much of that pile would have been cleared first. After that, academic prices are $149 per review (Replay, title/abstract) or $249 (Plus, which adds an AI full-text pass). Team is $29 per seat per month. Commercial plans are listed on the pricing page.
We'd love your feedback, especially from people who screen for a living:
- If you already use Covidence, would this pre-screen-then-export flow help?
- Where would you need more proof before trusting an auto-exclusion?
We'll be here all day answering questions.
Warm regards,
the SyRev Team