Oncology trial recruitment
Matches patients to biomarker- and stage-specific protocols where eligibility is dense.
Clinical Intelligence Agent
AI clinical trial matching
Coordinators screen charts by hand, one protocol at a time, and eligible patients are missed. Agnotic Technologies builds AI clinical trial matching software that screens records against eligibility criteria and hands coordinators a ranked, explained shortlist.
Coordinator-reviewedFHIR R4Explainable criteriaAudit Logged
Trusted by global innovators
What is AI clinical trial matching software?
AI clinical trial matching software screens patient records against each trial's inclusion and exclusion criteria and returns a ranked shortlist with the evidence behind every criterion. Agnotic Technologies builds it for sponsors, CROs and research sites, reading structured FHIR data and free-text notes, with a coordinator confirming every candidate before outreach.
Trial enrollment is a research bottleneck. Coordinators screen charts by hand against dense inclusion and exclusion criteria, one patient and one protocol at a time.
The software does first-pass screening at scale. It encodes each trial's eligibility as structured criteria, reads the patient's record and returns a ranked met, unmet or unknown shortlist with links to the data. It screens; it does not enroll.
Criteria parsing
Criteria parsing turns each protocol's eligibility text into structured, coded criteria that the matcher checks against FHIR patient data and facts extracted from notes, with an audit trail of every decision.
What does the matching software do?
Trial eligibility screening criterion by criterion on structured and unstructured data: explainable and coordinator-gated.
Eligibility matching
Evaluates a patient against each inclusion and exclusion criterion and returns met / unmet / unknown per criterion, not an opaque yes-or-no.
FHIR ResearchStudy criteria
Represents trials as FHIR ResearchStudy with structured eligibility, so protocols are machine-checkable and reusable across sites, not trapped in a PDF.
NLP for unstructured notes
Extracts stage, biomarker status, ECOG performance, and prior therapy from clinical notes so criteria in narrative text are still evaluable.
Ranked, explained, gated
Ranks candidate trials per patient with the criteria met and supporting data; every match is a candidate a coordinator confirms before outreach.
Where it runs
Who uses patient recruitment AI?
Site feasibility
Estimates how many patients could qualify for a protocol before the study opens.
Cohort screening
Runs a new protocol's criteria across the population via bulk FHIR.
Coordinator pre-screening
Hands coordinators a ranked, explained shortlist so review starts from candidates, not raw charts.
What changes for coordinators?
Directional: depends on protocol complexity and data quality.
At scale
First-pass screening across the eligible population
Per-criterion
Every match shows met / unmet / unknown with evidence
Human-gated
Coordinator confirms eligibility before any outreach
Compliance-First Healthcare App Development Services Backed by Global Standards
01
HIPAA
Health Insurance Portability and Accountability Act
Protect PHI with privacy-first architecture, encrypted storage and transmission, strict access controls, and traceable audit logs.
02
GDPR
General Data Protection Regulation
Implement lawful consent flows, data minimization, retention controls, and secure processing for sensitive health data.
03
FHIR
Fast Healthcare Interoperability Resources
Enable standardized health data exchange across apps, care teams, and systems through standards-based FHIR APIs.
04
HL7
Health Level Seven International
Support enterprise-grade interoperability with HL7-based integrations for records, events, and clinical messaging workflows.
05
HITRUST
Health Information Trust Alliance
Align security programs to healthcare-specific control and risk management practices trusted by providers and ecosystem partners.
06
HITECH
Health Information Technology for Economic and Clinical Health Act
Design with breach notification readiness, digital record safeguards, and operational controls that support regulated care programs.
07
SaMD
FDA Software as a Medical Device
Plan software quality, traceability, and documentation pathways for products that may require SaMD review and submission.
08
MDR (EU)
Medical Device Regulation (European Union)
Prepare EU market-ready processes for risk classification, evidence tracking, and lifecycle governance under MDR expectations.
09
SAMHSA
Substance Abuse and Mental Health Services Administration
Apply confidentiality controls and consent-aware sharing models for behavioral and mental health data experiences.
Standards we build against
Which standards does trial matching use?
FHIRBulk FHIRHIPAAGCPSNOMED
How It Works
From protocol to a reviewed shortlist
Structured criteria and FHIR patient data in, per-criterion evaluation with guardrails, a coordinator-reviewed shortlist, and a logged screening record.
1.
Capture: criteria + FHIR data
Trial eligibility is encoded as structured criteria (FHIR ResearchStudy), and the agent reads the patient's FHIR record plus NLP-extracted facts from notes.
FHIR R4 + NLP
2.
Processing: evaluate + guardrails
Each criterion is evaluated met / unmet / unknown; unknowns are surfaced honestly rather than assumed, and low-confidence NLP extractions are flagged for review.
Explainable matching
3.
Action: coordinator-reviewed shortlist
A ranked shortlist is presented with per-criterion evidence; a coordinator or investigator confirms eligibility and consent: the agent proposes, the human enrolls.
Coordinator-in-the-loop
4.
Write-back: screening record + audit
The screening result, matched criteria, and coordinator decision are recorded and logged to an immutable audit trail for regulatory and monitoring needs.
Audit logged
Facts and sources
Which rules shape trial matching?
Short facts we build trial matching around, each linked to its primary source.
In HL7 FHIR R4, the ResearchStudy resource describes a study's purpose, sponsor, condition and schedule of activities, and one listed use is registering a clinical trial in a registry such as ClinicalTrials.gov.
Source: HL7 FHIR R4, ResearchStudyThe ClinicalTrials.gov REST API (version 2.0) returns study records through its studies endpoint, and each record's eligibility module carries the study's eligibility criteria text.
Source: ClinicalTrials.gov APIThe HIPAA Privacy Rule lets covered entities use or disclose health information for research with a waiver of authorization approved by an IRB or privacy board.
Source: 45 CFR 164.512(i) (eCFR)21 CFR Part 11 sets FDA's requirements for electronic records and electronic signatures, including secure, computer-generated audit trails.
Source: 21 CFR Part 11 (eCFR)
Related proof
Read Case StudyLera Health: compliant women's health platform
Related proof of compliant, AI-assisted delivery, not this exact agent. Lera Health shows how we build FHIR-native data pipelines and compliant AI workflows that turn clinical data into structured output.
Compliance & Guardrails
Screening that respects consent and the record
Research touches PHI, consent, and regulatory scrutiny. The guardrails are non-negotiable.
PHI handling
Screening runs inside your compliance boundary under a BAA, encrypted throughout, scoped to authorized research staff and IRB-aligned data use.
Coordinator-in-the-loop
The agent produces candidates, not enrollments. A coordinator confirms eligibility and obtains consent before any patient is contacted.
Explainability
Every match shows which criteria are met, unmet, or unknown and the data behind each.
Audit logging
Screening decisions, matched criteria, and coordinator actions are logged immutably: the record trial monitoring and audits require.
Frequently Asked Questions
Screening runs inside your compliance boundary under a signed business associate agreement, encrypted throughout and scoped to authorized research staff, with every decision audit-logged. Whether records may be screened for recruitment, under authorization or an IRB waiver, is decided by your research team.
No. It produces a ranked, explained shortlist; a coordinator confirms eligibility and consent before any patient is contacted.
It evaluates a patient's FHIR record and NLP-extracted facts against structured FHIR ResearchStudy criteria, criterion by criterion, returning a ranked shortlist with evidence, not free-text chat.
Yes. NLP extracts details such as stage, biomarker status, ECOG performance status and prior therapies from notes and pathology reports, and links each fact to its source sentence. Low-confidence extractions are marked unknown for the coordinator rather than counted as met.
It depends on how many protocols you start with, how complex their criteria are, how much of the needed data sits in free text, which EHR or data warehouse we read from and how the shortlist fits the coordinator's workflow. We give you a timeline after the scoping call.