Early-stage preview. No active customers yet. Sharing progress as it's built.
🔒 Private AI · runs on your computer

From question
to publication.

The research platform that guides you through systematic reviews and meta-analysis, with AI that runs on your own machine, so your data stays there by default.

Request beta access
Free beta available now. Email us for access.
Local-first AI · no cloud LGPD & GDPR ready Built around gold-standard reporting guidelines
Local-first AI

Your research data stays on your computer.

By default, Compendium runs open AI models on your own hardware: no uploads of your research content. Ideal for ethics boards, LGPD/GDPR and institutions. There is no cloud option to opt into in the current beta: nothing leaves your machine by default.

Everything in one place

The whole research workflow, guided.

From protocol and screening through extraction, risk of bias, meta-analysis and PRISMA reporting, one guided pipeline, with local AI drafting and explaining itself at each step. Four moments below show what it feels like in the real app; every call stays yours.

I

Protocol

Guided PICO / PROSPERO. Pin down the question and lock a reproducible protocol, drafted by local AI, decided by you.

✦ Local AI draft
II

Search

Build a reproducible search strategy with synonym blocks and AND/OR logic, the raw per-database queries are generated live as you edit.

✦ PRISMA-S · reproducible
III

Judge the trust

RoB 2 flags trustworthiness across five domains: colour plus glyph, never colour alone. The tool proposes a judgment; you can override any domain.

✦ Five RoB 2 domains · human is final
IV

Report & write

A PRISMA diagram and the right reporting guideline are built from your decisions, with a methods draft, exported to PDF or DOCX.

✦ PRISMA · DOCX export

Screens shown in Portuguese: the app's interface (and this demo review) run in Portuguese today; an English UI pass is on the roadmap.

The real product

What you'll actually open.

Below, an actual screenshot of the desktop app: full-text screening, with the local AI's own suggestion, a leaning verdict and a literal quote from the PDF, offset-verified, right next to your decision.

Compendium· Full-text screening
Compendium's full-text screening screen with the local AI reader's suggestion panel open: a leaning verdict, a confidence bar, and a literal quote from the attached PDF

A real screenshot of the app's own local AI reader, explore it yourself once you're in. Design continues to evolve.

Built for all of research

Not just meta-analysis.

Compendium starts with systematic reviews and meta-analysis, built around gold-standard reporting guidelines. More study types are on the roadmap.

PRISMAsystematic reviewsLive now
+ more study types on the roadmap
Clinical trialsRoadmap

Randomized and interventional study designs

Observational studiesRoadmap

Cohort, case-control and cross-sectional designs

Case reportsRoadmap

Structured reporting of individual clinical cases

Diagnostic accuracyRoadmap

Studies that evaluate tests and diagnostic tools

Study protocolsRoadmap

Planning and registering research before it starts

Prediction & AI modelsRoadmap

Development and validation of predictive models

Economic evaluationsRoadmap

Cost-effectiveness and health-economics studies

Qualitative researchRoadmap

Interview, focus-group and mixed-methods studies

Each will be built around the gold-standard reporting guideline for its design, with the same guided, privacy-first, human-in-the-loop approach. None of these are available yet. Email us to help decide the order.

Systematic reviews & meta-analysis are available now; the rest are on our roadmap.
The result

Evidence, composed.

Compendium guides you to a submission-ready forest plot and a PRISMA diagram, with every number traceable to its source. A finished synthesis, not a pile of tabs.

Compendium's forest plot for six included studies, mean difference in sleep onset latency, pooled effect -5.50 minutes, 95% CI -7.17 to -3.84, I-squared 0%

Frequently asked questions.

Does the AI replace the researcher?

No. You stay the reviewer of record. The AI drafts, screens and shows reviewer agreement (κ), but every call is yours, with transparent AI-use disclosure.

Where does my data go, and who owns it?

By default, nothing leaves your machine: open local models only, no cloud option in the current beta. Ideal for sensitive data, ethics boards and LGPD/GDPR. Export to PDF/DOCX anytime, no lock-in.

How is this different from other systematic review tools?

Compendium runs the whole PRISMA pipeline in one place, screening, RoB 2, meta-analysis, PRISMA diagram, locally with AI. It's early access, with the core workflow real and usable today.

Is it only for systematic reviews?

Today: systematic reviews and meta-analysis (PRISMA), end to end. More study types are on the roadmap.

How mature is this, can I rely on it for real work?

We're upfront: early access, no paying customers yet. The core workflow is real and usable, but you'd be an early adopter, not buying a finished platform.

Is it good for beginners?

Yes. Built for your first study, with guided steps so PRISMA and RoB 2 come built into the flow, not something you assemble yourself.

Is Compendium available where I am?

For researchers in Brazil, the US and the UK, yes: a July 2026 legal review found no blocker for the current free beta in any of the three. The beta is local-first, runs without a backend and, by default, collects no personal data beyond an optional support email. That's not a compliance certificate, just no blocker found so far. Somewhere else? Email us and we'll look at your case.

Be among the first researchers.

Early access and the launch price for the first researchers who reach out. Help shape the roadmap.

Contact us
Help build Compendium.