Gary Riggs, MD — Practicing Physician & AI Developer

Healthcare AI That's
Actually Built for Medicine

Most "healthcare AI" tools are just general chatbots behind a login.
RiggsMedAI is different. These tools are custom-built medical models, designed by a practicing physician, to handle real clinical work — fast, predictable, and without relying on a general-purpose chatbot to do the core clinical work.

Built for physicians, medical directors, and healthcare teams who want AI that behaves predictably.

Why RiggsMedAI Exists

Healthcare doesn't need another chatbot guessing at medical decisions.
RiggsMedAI tools are designed to do one clinical job at a time — billing, de-identification, analytics — using purpose-built AI models, not general-purpose language models trained on the internet.

🎯

No speculative
"AI advice"

🔄

No prompt
guessing

No unpredictable
behavior

💵

No surprise
usage costs

⚡ Built for Healthcare — Not Powered by a Chatbot

Most AI healthcare products work like this:
Your clinical text is sent to a large chatbot, and the response comes back seconds later.

RiggsMedAI works differently.
These tools use custom-trained deep learning models, built specifically for medical documentation and workflows.

Instant responses (milliseconds, not seconds)
🔒No clinical data sent to public chatbots
💵Predictable costs — no per-word billing
🧠Models trained on medical structure

Some tools may use large language models selectively for explanation or review — never as the primary decision engine.

This isn't "AI chat." It's clinical-grade automation.

Tools That Actually Work

🧠

CPT Coding Assistant

Reads a physician note and suggests the correct E/M code — instantly.

This is not a chatbot guessing at billing. A language model is used only for post-hoc analysis and explanation — not for code prediction. It's a custom deep neural network, trained on real hospitalist documentation to predict E/M codes with clinical-grade accuracy. See exactly which parts of the note influenced the code.

96%+
Accuracy
~50ms
Inference
$0
Per Query
Try it live
🔒

Clinical De-identification

Removes PHI while keeping clinical meaning intact.

Instead of replacing names with ugly [REDACTED] blocks, this tool generates realistic surrogates that preserve readability and context. Built using a fine-tuned clinical language model, not a generic text filter. Open source on Hugging Face.

97.7%
F1 Score
~100ms
Inference
Free
Open Source
Try it live
📊

HCAHPS Analytics

Real-time insight into physician performance and patient experience.

This system tracks satisfaction metrics across multiple facilities and delivers live dashboards, automated trend analysis, and SMS alerts for score changes. Designed for medical directors who want visibility without spreadsheet chaos.

3
Facilities
Live
Updates
SMS
Alerts
View Dashboard
🔍

ChargeScrub

Find duplicate and overlapping charges before they become denials.

Upload a weekly charge report and instantly identify duplicate E/M charges, overlapping encounters, and visits that need timeline review. Built with HIPAA-aligned workflows and zero long-term data storage.

HIPAA
Aligned
Zero
Storage
Instant
Results
Try it live
👨‍⚕️

Physician-Built AI

I'm Gary Riggs — a practicing hospitalist and Medical Director.

I write notes. I deal with billing headaches. I answer patient complaints. And I see firsthand where technology helps — and where it fails.

RiggsMedAI tools are not academic experiments or investor demos. They're built to survive the messy reality of real clinical work.

  • Medical Director, Metro Physician Group
  • Hospitalist at SSM Health (OKC, Midwest City, Shawnee)
  • MS in Data Science, Northwestern University
  • Founder, RIGGSMED LLC
PyTorch AWS Transformers FastAPI Serverless

Built With Healthcare Reality in Mind

No clinical data used to train public AI models

No prompt logging to third-party chatbots

Narrow, auditable models — not open-ended AI

Built and maintained by a practicing physician

Let's Talk

Interested in tools that prioritize accuracy, speed, and accountability over hype?

riggsmed@gmail.com