Systematic literature review software
Most tools only help you screen. AutoSearch runs the whole systematic review — search, screening, synthesis, PRISMA flow and verified citations — from one research question to a defensible, IMRAD-structured paper.
Built in Switzerland · Crossref-verified citations · 0 fabricated DOIs in our public benchmark
A PRISMA 2020 systematic review is slow: define the protocol, search multiple databases, screen hundreds of records, extract and synthesise, document everything. Screening tools (Rayyan, Covidence) help with one step; generalist AI (ChatGPT) invents citations. AutoSearch does the full pipeline with real, verified sources — so researchers and PhD students get a defensible review in an afternoon, not months.
The whole review, not just screening
AutoSearch searches OpenAlex, Crossref, PubMed, Semantic Scholar, arXiv and ClinicalTrials.gov in parallel, scores and screens each record with a recorded decision, synthesises the evidence, and outputs a PRISMA 2020 flow diagram, a typed evidence matrix, inline citations, a BibTeX file and an exportable screening log — plus an IMRAD manuscript and a LaTeX/PDF preprint.
Verified citations, zero hallucinations
Generalist AI fabricates 3–18% of citations. AutoSearch verifies every DOI live against Crossref — our public benchmark reports 0 fabricated DOIs across hundreds of references. Your review rests on sources a reviewer can actually open.
Reproducible and exportable
The exact search protocol is stored and re-runnable, the screening log exports to CSV, and references export to BibTeX — everything you need for a transparent methods section and a PRISMA-compliant submission.
| Review stage | AutoSearch |
|---|---|
| Protocol & search question | Structured objective captured and documented; reproducible. |
| Multi-database search | OpenAlex, Crossref, PubMed, Semantic Scholar, arXiv, ClinicalTrials.gov. |
| Screening | Recorded keep/discard decision per record; CSV log. |
| Synthesis & appraisal | Typed evidence matrix with confidence; IMRAD manuscript. |
| PRISMA 2020 flow | Flow diagram generated automatically with counts. |
| Export | BibTeX, PRISMA, screening log, LaTeX/PDF preprint. |
Frequently asked questions
How is this different from Rayyan or Covidence?
Those are excellent screening tools — they help you de-duplicate and screen records you import. AutoSearch covers the whole pipeline: it does the multi-database search, screening, synthesis, PRISMA flow and a drafted IMRAD review with verified citations.
How is it different from ChatGPT for a literature review?
Generalist chatbots fabricate citations (3–18% in studies). AutoSearch verifies every DOI live against Crossref — 0 fabricated in our public benchmark — and documents a reproducible search protocol.
Is it PRISMA 2020 compliant?
It generates a PRISMA 2020 flow diagram, an evidence matrix and a documented, reproducible search — the transparency PRISMA expects. You retain authorship and final judgement.
Which databases does it search?
OpenAlex, Crossref, PubMed, Semantic Scholar, arXiv, ClinicalTrials.gov, DBLP and DOAJ, in parallel.
How much does it cost?
From CHF 29/month (Researcher, 300 credits) or one-time credits; a standard review is about 8 credits, with 50 free starter credits on signup — a fraction of enterprise SR suites.
Run your systematic literature review now
Write one research question and AutoSearch returns a PRISMA-grade, DOI-verified systematic review with an IMRAD manuscript and exportable BibTeX.
Run a free review →