For institutions

For universities, research institutes, and funders

Run pre-submission review at scale. Sponsor open-access, transparent peer review.

3
Supported workflows
10+
Disciplines
8
Agents per paper
CC BY 4.0
Open access

Three institutional workflows

Most institutional conversations we have fall into one of three patterns. We can support all three with the same underlying pipeline; the integration depth differs.

  • Pre-submission QA gate: every paper leaving the institution runs through the 8-agent pipeline before external submission. Aggregate scores surface systemic issues (statistics gaps, reproducibility weaknesses) for targeted training.
  • Preprint-server QA layer: institutional preprint servers integrate our pipeline as an optional 'AI-reviewed' badge on uploaded preprints, visible alongside the abstract.
  • Open-access publication venue: institutions fund publication of their researchers' papers in Science AI Journal, bypassing closed-journal APCs and subscription paywalls. All papers remain CC BY 4.0, fully open access.

What institutional sponsorship funds

We are an open-access venue with no APCs and no paywalls. Institutional sponsorship is the sustaining model for this posture — and it is the fair one.

  • Compute for the 8-agent pipeline (Anthropic Claude + Ollama fallback).
  • Curation and calibration of the 23,000-review training corpus.
  • Ongoing prior-publication index maintenance (900K papers + 6 external databases).
  • Editorial advisory board honorariums.
  • Platform infrastructure, security, and long-term archival commitments.

Data posture

We are boring about data. Manuscripts uploaded for institutional review are stored encrypted, never added to the training corpus, and deletable on institutional request at any time. Reports are accessible to the author and to whichever institutional administrators the author designates — nobody else.

We are GDPR-ready, KVKK-ready (Turkish equivalent), and working toward SOC 2 Type II. Full details on /policies.

How to start a conversation

If you represent a university, research institute, funder, or preprint server and want to discuss any of the three workflows above, email [email protected] with a short note. We will respond within one business day with a concrete proposal sized to your institution.

We do not run outbound sales calls or quarterly business reviews. We do answer email.

Frequently asked questions

Yes — the Ollama fallback path supports full on-premises deployment for institutions with data-residency requirements. Performance is somewhat lower than the Claude path but the rubric and pipeline are identical.
Email editorialRead the engineering blog

Command palette

Jump anywhere, run any action.