Vol. I · MMXXVIEvidence-grade clinical AI
DiscensMachinaEvidence-grade clinical AI
Request demo
Product 01 · Live at cliniatlas.com

AI answers tomedical questions,grounded in evidence.

Ask a clinical or research question. CliniAtlas searches PubMed, Europe PMC, clinical guidelines, and regulatory sources, then synthesises a cited answer — with evidence quality and links you can inspect.

Live product

CliniAtlas is live — two trial questions, no sign-in required.

Open CliniAtlas
Overview§ 01
Ask a question. Get a cited answer —graded by the evidence.

CliniAtlas searches peer-reviewed literature, European guidelines, and the EMA medicines register, then writes an answer in which every claim carries its source. Sources are placed by their stratum in the evidence hierarchy — not by popularity.

The interface
Clinical research question

What are the benefits and harms of melatonin for insomnia in adults?

No sign-in requiredAsk CliniAtlas
Sample readings

Grounded synthesis · Citations checked

The reading

Cited answer, graded by source.

Example readingGrounded synthesisCitations checkedIllustrative reading

What cardiovascular and kidney benefits, harms, and uncertainties are reported for GLP-1 receptor agonists in adults with type 2 diabetes?

A meta-analysis of eight placebo-controlled cardiovascular outcome trials involving 60,080 participants found that GLP-1 receptor agonists reduced major adverse cardiovascular events (MACE) by 14% (HR 0.86, 95% CI 0.80–0.93). It also reported relative reductions in all-cause mortality, heart-failure admission, and a composite kidney outcome. [1]

In LEADER, liraglutide reduced the primary MACE outcome from 14.9% to 13.0% over a median 3.8 years (HR 0.87, 95% CI 0.78–0.97). REWIND enrolled a broader-risk population: dulaglutide reduced MACE from 13.4% to 12.0% over 5.4 years (HR 0.88, 95% CI 0.79–0.99). [2][3]

SUSTAIN-6 reported fewer MACE with semaglutide, but more diabetic-retinopathy complications (3.0% versus 1.8%; HR 1.76, 95% CI 1.11–2.78) — an agent-specific safety signal that should not be hidden inside a class-average benefit estimate. [4]

The 2023 ESC guideline incorporates evidence from large cardiovascular outcome trials into risk-focused management. Guideline context helps translate trial evidence, but it does not make every GLP-1 receptor agonist or every patient population interchangeable. [5]

Limitations
  • Most cardiovascular outcome trials enrolled people with established cardiovascular disease or elevated risk; absolute benefit is likely smaller in lower-risk populations.
  • Gastrointestinal intolerance is common, and the retinopathy signal in SUSTAIN-6 requires clinical context rather than simple class-wide extrapolation.
  • The landmark trials were largely industry funded and generally compared treatment with placebo on top of standard care.
Sources in this reading
  1. [1]III / Systematic review2021

    Cardiovascular, mortality, and kidney outcomes with GLP-1 receptor agonists in type 2 diabetes: a systematic review and meta-analysis of randomised trials

    The Lancet Diabetes & Endocrinology

    PMID 34425083
  2. [2]IV / Randomised trial2016

    Liraglutide and cardiovascular outcomes in type 2 diabetes (LEADER)

    New England Journal of Medicine

    PMID 27295427
  3. [3]IV / Randomised trial2019

    Dulaglutide and cardiovascular outcomes in type 2 diabetes (REWIND)

    The Lancet

    PMID 31189511
  4. [4]IV / Randomised trial2016

    Semaglutide and cardiovascular outcomes in type 2 diabetes (SUSTAIN-6)

    New England Journal of Medicine

    PMID 27633186
  5. [5]I / Guideline2023

    ESC Guidelines on the management of cardiovascular disease in patients with diabetes

    European Heart Journal

    ESC 2023
Atlas sheets

What CliniAtlas reads.

3 source classes
GuidelinesIndexed · EU

European society & national

Curated guidance corpus — European, UK and society sources, versioned with issue dates.

LiteratureOn demand · EU

PubMed / Europe PMC

Agentic search across peer-reviewed literature, retrieved per question with a selective cache.

RegulatoryIndexed · EU

EMA medicines register

Drug labels and assessment reports — SmPC and EPAR documents with source and version.

How a reading is made

Method & boundaries.

Path A / Reference tool

Evidence review only

For literature search, education, and professional evidence review only. Not for diagnosis, treatment selection, triage, monitoring, prognosis, or patient-specific clinical decision-making.

Provenance

A citation on every claim

Each line of an answer maps to retrieved sources with an identifier and licence status. Claims that cannot be cited are held back.

Hierarchy

Graded, not popularity-ranked

Sources are placed by their stratum in the evidence hierarchy and recency — not by engagement or citation counts.

Boundary

Read and processed in the EU

Retrieval and synthesis run in eu-central-1. No patient identifiers are entered, and outputs remain the professional's judgement to verify.

Academic collaboration

Built with the University of Split School of Medicine.

CliniAtlas is created by Discens Machina, a startup spin-off from the University of Split School of Medicine, with the goal of making clinical evidence easier to inspect in European practice.

More about us →
University of Split School of Medicine

Every claim stays inspectable. Nothing to take on faith.

Next step

Ask your first question — no sign-in required.

CliniAtlas is live. The cited reading shows exactly how evidence is graded, linked, and left open for your inspection.

Try CliniAtlas
Contact us
hello@discensmachina.com