Snow Owl offers an AI-driven platform that extracts clinical concepts from electronic health records, pathology reports, and imaging notes, normalizes them to standard ontologies, and automatically matches patients to oncology and rare‑disease trial eligibility criteria. The solution integrates with hospital EHRs via FHIR, provides secure real‑time notifications to study coordinators, and includes an analytics dashboard for monitoring enrollment performance.
Funding
Funding not disclosed
Founders
Product
Problem
Clinical trial enrollment is often delayed because investigators must manually review heterogeneous, unstructured patient records to assess eligibility for oncology and rare‑disease studies. This manual process is time‑consuming, error‑prone, and limits the pool of patients who are matched to appropriate protocols.
Solution
Snow Owl delivers an AI‑driven matching platform that automatically extracts clinical concepts from electronic health records, pathology reports, and imaging notes using domain‑specific natural‑language processing. The extracted data are normalized to standard ontologies (e.g., SNOMED CT, MedDRA) and evaluated against trial inclusion/exclusion criteria encoded in a rule engine. Eligible patients are flagged in real time and routed to study coordinators through secure notifications. The system integrates with hospital EHRs via FHIR APIs, ensuring seamless data flow while maintaining HIPAA‑compliant encryption. Sponsors and sites can monitor enrollment metrics on a web‑based analytics dashboard, enabling proactive adjustments to recruitment strategies. By automating the eligibility assessment, the platform shortens the time from patient identification to consent, improving overall trial timelines.
Target Audience
Primary users are pharmaceutical sponsors, contract research organizations, and academic research sites that conduct oncology or rare‑disease clinical trials and need scalable patient‑matching capabilities.
Features
- Domain‑specific NLP pipeline that parses unstructured clinical notes, pathology reports, and radiology summaries into structured phenotype vectors
- Criteria‑matching engine that applies Boolean and probabilistic rules to map patient phenotypes against trial protocols
- FHIR‑based bidirectional integration with major EHR vendors for real‑time data ingestion and patient flagging
- Role‑based access control and end‑to‑end encryption to meet HIPAA and GDPR requirements
- Interactive recruitment dashboard with cohort analytics, enrollment funnels, and protocol‑level performance KPIs
- Automated outreach module that generates secure, consent‑ready communication templates for study coordinators
- Extensible RESTful API for sponsor CRO systems to pull matched patient cohorts and feed trial management platforms
- Support for oncology and rare‑disease ontologies, including cancer staging (TNM) and orphan disease registries