UPMC · Step By Step Hip & Knee · Healthcare

AI Clinical Reporting Platform

HIPAA-compliant clinical platform that uses AI to analyse hip and knee X-rays and turn findings into patient-friendly reports.

80%faster report generation
HIPAAcompliant end to end
DICOMX-rays analysed by AI
8 mokickoff to delivery
AI Clinical Reporting Platform for UPMC · Step By Step Hip & Knee
My role

Full-Stack Developer · React, Node.js/AdonisJS, Python, AWS

  • Built the clinician-facing React.js evaluation workflow (Context API) for X-ray review, chart rating and patient education resources.
  • Developed the Node.js / AdonisJS REST API and Lucid ORM data layer on AWS RDS.
  • Integrated the Python DICOM imaging services with the API and turned AI findings into automated, patient-friendly reports.
  • Designed a custom job queue on AWS EC2 and Lambda so heavy analysis jobs never block the clinician experience.
  • Implemented HIPAA-aligned security: TLS via AWS ACM, Nginx hardening, controlled access and secure storage.
The challenge

SBS (Step By Step Hip & Knee) is a clinical evaluation and treatment platform developed with UPMC for arthritis care. Clinicians needed one secure place to evaluate X-rays, rate patient charts and share educational resources.

The platform had to analyse DICOM X-ray images with AI, turn the results into reports patients could understand, meet HIPAA requirements throughout and stay fast under a high volume of requests.

The solution

A full-stack platform with AI imaging analysis at its core: a React.js front end for a clean clinician workflow, a Node.js and AdonisJS API, and Python services for the imaging work, all on hardened AWS infrastructure.

01

AI X-ray analysis

Python services process DICOM images and feed structured findings into the clinician workflow.

02

Patient-friendly reports

Findings become clear reports automatically, replacing hours of manual write-up.

03

HIPAA-compliant data

Encrypted transport with AWS ACM, controlled access and secure storage on AWS RDS.

04

High-volume queue

A custom queue on EC2 and Lambda absorbs spikes so analysis jobs never block the app.

The results
  • Report generation became 80% faster, cutting manual preparation time significantly.
  • Delivered a HIPAA-compliant AI medical evaluation platform.
  • Arthritis evaluation workflows became more efficient for clinicians.
Next case study

Multilingual Sites & Custom CMS →

Verizon Wireless · Fios · 2 Verizon brands localised

Let's talk

Need something similar?

I build web apps, mobile apps and AI automation end to end. Tell me what you're working on.