Literature review where every claim is cited.

One query across the whole record.

  • Search papers, patents and protocols at once.
  • Every sentence in a result links to the passage it came from.
  • Open questions and conflicting findings are marked, not hidden.

One query over papers, patents and protocols · three become citations

Three agents, one result.

You ask · agents take turns · results arrive cited · you sign off

  • Reader long-context LLM

    Reads full texts and pulls out the claims that answer your question.

  • Analyst code model

    Compares findings across studies and flags conflicts.

  • Writer drafting LLM

    Drafts the review section by section, citing as it writes.

What you get back.

Evidence table

Each claim with its source, the study type and the passage.

Review draft

A structured draft with numbered citations you can open.

Open questions

Gaps and conflicts listed for your team to judge.

Citation export

Sources travel with the text into your report.

Common questions.

Which sources can the agents search?

Published papers, patents, protocols and registries, plus the documents you bring into your project. Each result shows where it came from, so you can open the original and read the passage yourself before you rely on it.

How do I check a citation?

Every claim carries a numbered marker. Open it and you see the source and the passage the claim is based on. Claims the agents could not link to a passage are marked as open questions, so they stand out instead of blending in.

Can I add my own papers and notes?

Yes. Upload or connect them to your project. They are searched together with everything else, stay inside the project and never train shared models. Only the people you add to the project can see them.

Bring your next question.

Tell us your field and what you want to do. We set up the workspace around it.

Request access