Skip to content

srikanth2328/Predictive-Medical-Billing-System-Using-Azure-and-AI

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

2 Commits
 
 

Repository files navigation

Predictive-Medical-Billing-System-Using-Azure-and-AI

This project uses Artificial Intelligence (AI), Natural Language Processing (NLP), and Machine Learning (ML) to automate and optimize the medical billing and claims approval process. Built using Python, Azure Machine Learning, and Power BI, this system identifies patterns in healthcare claims, predicts approvals, and flags potentially fraudulent claims for auditing.


Project Overview

Objective:
To build a scalable, AI-powered medical billing analysis system that can:

  • Classify and analyze claims data
  • Predict claim approval status
  • Detect anomalies and fraud
  • Provide dashboards for real-time insights

My Tools & Used Technologies

  • Python: Data processing, modeling
  • Pandas / NumPy / Scikit-learn / XGBoost / IsolationForest
  • NLP: SpaCy, BERT (for analyzing claim notes)
  • Azure: Machine Learning Studio, Blob Storage, Form Recognizer
  • Power BI / Streamlit: Dashboard and visualization
  • SQL: Data exploration and preprocessing

Dataset Sources

  1. CMS Medicare Claims Data

Workflow

  1. Data Ingestion: Claims data collected from public datasets or synthetically generated
  2. Preprocessing: Cleaning, encoding, normalization, and NLP on notes
  3. Modeling:
    • Approval Prediction Model (Logistic Regression, XGBoost)
    • Fraud Detection Model (Isolation Forest, Autoencoder)
  4. Deployment: Models deployed using Azure ML as REST APIs
  5. Visualization: Dashboards built in Power BI for actionable insights

Sample Use Cases

  • Hospitals auditing claims for risk and fraud
  • Insurance companies prioritizing claims processing
  • Revenue cycle teams identifying high-risk claims proactively

Further work Integrations

  • Integrate OCR with Azure Form Recognizer for scanned forms
  • Add real-time streaming with Azure Stream Analytics
  • Expand to ICD-10/CPT hierarchy analysis and explainable AI

Contact

Built with 💙 by [Srikaanth Pavuluri]
Feel free to contribute or open an issue!


Credits

Datasets or data used for this project is owened by its respectrive owners. Thanks for your understanding

About

Personnel project on training an AI on medical bill claims processing

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published