Retinex Healthcare EHR
Healthcare software featuring AI-assisted medical coding and billing modifier recommendations for clinical workflows.
- Python
- Node.js
- AI/ML
- Healthcare APIs
Schematic illustration. Product screenshots are not shown to protect confidential information.
Overview
Retinex is a healthcare EHR platform developed at Auxiliobits. I contribute to its AI-assisted medical coding workflows, which recommend codes and billing modifiers to support clinical and billing teams.
The Problem
Medical coding and choosing billing modifiers are detailed, rules-heavy tasks. Supporting coders with relevant recommendations inside their existing workflow can reduce repetitive effort.
Architecture
Clinical data
- Encounter information
- Healthcare APIs
Services
- Python services
- Node.js services
AI assistance
- Medical code recommendations
- Billing modifier recommendations
Users
- Review within the EHR
- Billing workflow
How It Works
- 1
Clinical data
Encounter information is made available through healthcare APIs.
- 2
AI assistance
AI/ML services generate coding and billing modifier recommendations.
- 3
Review
Recommendations are surfaced within the EHR for users to review.
- 4
Billing
Confirmed codes and modifiers continue into the billing workflow.
My Contribution
- Contributed to the Retinex Healthcare EHR and its AI-assisted medical coding workflows.
- Built and maintained backend services with Python and Node.js.
- Integrated APIs, databases, and cloud services.
- Took part in testing, debugging, deployment, and production support.
Engineering Challenges
- Fitting AI recommendations into established clinical and billing workflows instead of adding friction.
- Working carefully with sensitive healthcare data and integrations.
Lessons Learned
- In regulated domains, AI works best as assistance that people can verify, not as an opaque decision.
- Healthcare integrations reward defensive engineering and thorough testing.