Industries · Healthcare & Life Sciences
AI for healthcare operations and research
Clinical language is our home turf. We build AI that reads, codes and validates clinical information for insurers, hospitals and pharmaceutical companies, with the rigor that regulated healthcare data demands. We also take on the operational problems around it, project by project.
Research
Synthetic patients · PALS
Clinical AI cannot be built by experimenting on real patient data. PALS (PAtients Like uS) is our research line on synthetic patients: medically coherent, patient-level data and health records that are realistic yet not real, so clinical AI can be trained and validated without touching a single real record.
The approach rests on five principles: representativeness, utility, robustness, privacy preservation and transparency, aimed at making synthetic patients acceptable for regulatory submissions.
Talk to us about synthetic patientsPeer-reviewed · Frontiers in Digital Health, 2026
A Call for Principle-Based Acceptance of Synthetic Patients: Establishing the AI-Driven Foundation for Regulatory Submissions
Read the paperOperational AI
Where we work today
Three fields where we already have work in production.
Medical coding
Automatic mapping of clinical terms to standardized codes (ICD-10-CM, CPT, SNOMED): consistent, auditable, with minimal manual intervention. The same engine that codes claims inside Claims47.
Clinical document intelligence
Handwritten health declarations, medical reports and clinical forms read, structured and semantically analyzed, with medical advisors keeping the expert decisions.
Administrative automation
The document-heavy back office of healthcare: HR contract generation for hospitals, regulatory data orchestration for pharma market access, invoicing workflows.
In production
Who we do it with
Mutual Médica
Handwritten health declarations analyzed with AI, reducing human workload by up to 80% with final decisions in expert hands.
Hospital Sant Joan de Déu
Temporary employment contracts generated automatically between planning and payroll, saving up to 50 hours of manual work per month.
Boehringer Ingelheim
Five regulatory sources orchestrated into one enriched record per medicine for the Market Access team.
Working with clinical data?
Tell us about your project. We are happy to share how we handle clinical language, compliance and evaluation.