Nietzsche
An AI-assisted healthcare platform that turns complex referral documents into structured patient data.
Khrystyna Viitov
BSA at Trinetix
A healthcare platform designed to make complex dialysis referrals faster and easier to process.
Dialysis referrals often arrive as large packets of unstructured medical documents that require extensive manual review. I redesigned the intake workflow and introduced AI-assisted extraction to turn this information into structured patient data.
The experience combines AI automation with human verification, allowing nurses to focus on uncertain cases instead of reviewing every document manually.
Challenge
How do you introduce AI into a clinical workflow without taking control away from the people using it?
Nurses needed to process large referral packets while maintaining accuracy and clinical accountability. The existing CRM also relied on legacy UI patterns and Kendo UI components, limiting how much the experience could evolve.
The challenge was to reduce manual work while keeping every AI decision visible, understandable, and easy to verify.
Solution
I designed an AI-assisted workflow that turns document-heavy intake into a structured review process.
The system extracts information from incoming documents, organizes it into meaningful categories, and highlights AI-generated values with confidence indicators. Nurses can verify uncertain information, compare alternatives, or manually review complex cases.
I also redesigned the core CRM workflows and established consistent interaction patterns within the existing technical constraints.
Result
A faster intake experience that lets nurses focus on exceptions instead of routine document processing.
The new workflow significantly reduces manual scanning and data entry during intake. AI handles around 80% of cases autonomously, while nurses remain in control of low-confidence and complex referrals.
The redesigned experience also creates a stronger foundation for expanding AI capabilities across the product.
Next
Project.
Each project shows how I approach design, structure, and development.
