CliniAtlasClinical evidence atlas / EUCliniAtlas searches PubMed and other medical research databases, then stratifies the evidence by clinical relevance, methodological strength, and research impact. Every answer is built for source-level inspection, not black-box summarization.
For literature search, education, and professional evidence review only. Not for diagnosis, treatment selection, triage, monitoring, prognosis, or patient-specific clinical decision-making.
The product direction is shaped around European clinical practice, transparent source inspection, and evidence grading. The collaboration with the University of Split School of Medicine helps keep the product close to real teaching, research, and clinical review workflows.
CliniAtlas was created by DiscensMachina, a startup spin-off from the University of Split School of Medicine. DiscensMachina builds clinical AI systems that keep users in control of judgement, provenance, and data boundaries.
For literature search, education, and professional evidence review only. Not for diagnosis, treatment selection, triage, monitoring, prognosis, or patient-specific clinical decision-making.
CliniAtlas is built around verifiable evidence trails: each answer keeps its citations, source strata, journal metadata, abstracts, snippets, and links close at hand so users can inspect the reasoning before relying on a summary.
CliniAtlas works with medical institutions that want clinical evidence to be easier to find and to check.
CliniAtlas is built in collaboration with the University of Split School of Medicine (MEFST). DiscensMachina, the company behind CliniAtlas, is a spin-off of the school.
Since September 2026, every member of the Croatian Oncological Society has free access to CliniAtlas. Members sign in with their Google or Apple account.
Read the Society's announcement (in Croatian)CliniAtlas has three parts: Chat, for answering clinical and research questions from the literature; CliniAtlas Learn, for studying; and CliniAtlas Agent, for multi-step work in hospital teams. Inside the chat, pick the mode that matches how much rigor the question deserves.
Ask a clinical or research question and get a cited answer grounded in PubMed, Europe PMC, clinical guidelines, and regulatory sources. Every answer keeps its evidence trail open so you can inspect the sources behind each claim.
Scans the most relevant abstracts and answers in seconds. Best for quick fact checks and first orientation on a topic.
Searches more sources, reads open full texts where available, and reranks the evidence before answering. Best for questions where coverage and nuance matter.
An agent workflow on top of Full mode: it plans the search, drafts an answer, reviews the draft against the retrieved sources, and revises it before you see it. Slower, with the strongest quality control.
A longer running job that investigates a question from several angles and returns a structured, citation rich report you can save and export. Best for topics that deserve a proper write up.
A separate study conversation for medical students and clinicians. Ask about any concept at beginner, student or expert level, then follow up naturally: summarize, go deeper, compare. Quizzes, flashcards and Feynman breakdowns are built only when you ask, grounded in open textbooks and curated reviews. Create a workspace, add your own lecture files, and study from your materials with every answer labeled by its source.
