Showcasing some of my technical projects. All work shown is non-proprietary (which is why I can't show you everything).
What was implemented is a 2025 de-identification AI from Azure Health Data Services to remove PHI from patient data. HOWEVER, the example below is an OpenAI privacy-filter that pretty much works the same way, except I don't charged for the Azure AI tokens I use. I still recommend the Azure Health Data Services version as it specifically targets PHI.
Tech: Flask Python, Azure, PHI, Nginx
This is a 2025 implementation of OpenAI's Whisper transcription model that was used to add subtitles and captions to hundreds of patient-facing videos, hence the legal requirement for multi-language captioning. Open AI's model doesn't translate, but it's small task to simply push the subtitle files through a translator.
Tech: Flask Python, OpenAI, PHI, Nginx
Unlike the AI projects above, I did not perform any of the software development work on this 2021 product. MPD is a mobile and web app that enables clinicians (nurses and physicians) to message each other without relying on SMS. API hooks between the app's back-end and EMR allow clinicians to view their assigned patient charts.
Tech: SQL Server, Azure, Node.js, GraphQL
So I can't show anything about this project, as it's 100% proprietary, but essentially our developer processed HL7 EDI files and converted them to hierarchical JSON files. I created a script to import the JSON files into MongoDB and viola, we have a DB that we can then use to match patient claims to insurance remittances (until someone changes the EDI format, at least).
Tech: MongoDB, Python, Node.js, EDI 835 remittance, EDI 837 claim
Email: rob@thaxra.com
Location: Irvine, CA