Skip to content

AI drafts, the investigator decides

From protocol to a candidate list — the investigator decides

ClinicalTrials structures protocol criteria, finds candidates on site data without leaving its contour, and keeps study documents with signatures and an audit trail.

Looking for a trial for yourself? Find a trial

Pre-screening · DNA-CARD-204Demo data

AI criterion draft · protocol §4.1.4

NT-proBNP ≥ 300 pg/mL within 90 days

Observation(NT-proBNP) ≥ 300 · age ≤ 90 d

KY-0307

Meets all criteria

KY-0182

NT-proBNP missing — test needed

The investigator decidesInvite
Demo data. Real patient data never leaves the site contour.

Who it is for

Sites find patients, sponsors see the network picture

For research sites

Coordinators and investigators

  • Protocol criteria in minutes — the coordinator only reviews and approves
  • Candidates from site data with a reason for every criterion
  • Visits, patient requests, eTMF-lite and AE reports in one place

For sponsors and CROs

Medical monitors, study managers

  • Network feasibility — aggregates only, no personal data
  • Enrolment against plan by protocol and site
  • Investigator-signed safety reports and audit trail extracts

How it works

A coordinator’s day in four steps

  1. 1

    Upload the protocol

    AI proposes inclusion and exclusion criteria citing the protocol section.

  2. 2

    The coordinator approves

    Each criterion is accepted, edited or rejected. Rules are evaluated by code, not a model.

  3. 3

    The investigator picks candidates

    Pre-screening runs on site data inside its contour; every criterion shows its reason.

  4. 4

    Documents and safety

    eTMF-lite with versions and signatures, AE report drafts for the investigator to sign.

Features

What AI does and what people do

AI never diagnoses or decides: it prepares sourced drafts, and a specialist confirms every decision.

  • Protocol → structured criteriaAI: Extracts criteria, structures rules, cites the sectionPerson: The coordinator reviews and approves each
  • Pre-screening in the site contourAI: Not involved: approved rules run as codePerson: The investigator decides on each candidate
  • Registry searchAI: Simplifies ClinicalTrials.gov, EU CTIS, SEC records in UkrainianPerson: The coordinator reviews before publishing
  • Feasibility for the sponsorAI: Not usedPerson: Sites share aggregates only; cells < 5 hidden
  • eTMF-liteAI: Not usedPerson: Versions, signature with meaning and re-authentication
  • Adverse event reportsAI: Fills the sponsor template with sourcesPerson: The investigator assesses causality and signs

Security & compliance

Built on D.N.A. Health Core and GCP requirements

D.N.A. Health Core

  • Data in Ukraine / EU, DCP field encryption
  • Immutable hash-chained audit trail
  • Granular consent; no consent — no processing
  • Specialist in the loop: AI output is a draft until confirmed

GCP compliance

ICH-GCP E6(R3)
Site and sponsor processes
21 CFR Part 11 · EU Annex 11
E-signature, audit trail, version control
CSV validation: IQ / OQ / PQ
Validation package — on request, after the MVP
GDPR · EU AI Act
Separate health-data consent, AI log

Pilot & case study

The pilot is running — join the next wave

What we do

We parse 1–2 of your protocols and run pre-screening on site data.

What we measure

Criteria extraction completeness (target ≥ 95 %) and coordinator time before and after.

What we publish

Figures only after the pilot and with the site’s consent. Until then there are none here.

Pricing

A simple site plan, a contract for sponsors

Draft pricing. No online payment: request → invoice → contract.

Research site

$490 / mo

Protocols, pre-screening, patient requests, eTMF-lite, AE reports, audit trail.

Connect a site

Sponsor / CRO

Contract

Per study: network feasibility, enrolment, safety, reporting.

Discuss a study

FAQ

Frequently asked questions

Where is patient data stored?

Inside the site contour in Ukraine or the EU, health fields encrypted with the organisation’s key. Sponsors see aggregates only.

Who is responsible for inclusion decisions?

The investigator. AI only drafts criteria, eligibility is computed by deterministic code, and every decision is signed by a person.

Does the AI train on our site data?

No. Training on health data is off by default and possible only de-identified, under contract and with patient consent.

How do we connect?

Leave a request, we walk through your protocol together, and your admin invites the team into D.N.A. ID — one account for all products.

Is there integration with the site’s EHR?

Yes, via the Health Core FHIR R4 API. The specific EHR is agreed at the start of the pilot.

Is there a validation package for audits?

Yes, IQ / OQ / PQ for 21 CFR Part 11 and EU Annex 11 — available on request once the MVP is complete.

How do patients get into a study?

Through public search without an account: the patient leaves a request with separate consent and the coordinator calls back. No promises of treatment or reward.

Request

Let’s discuss your study

Tell us who you are and what you work on — we’ll demo the system on your criteria.