Medical Policy Analysis GenAI Prototype

Cheista Health - Chicago, IL

Oct 2023 - Dec 2023

Introduction

This tool is an advanced Generative AI-powered tool designed to automate the analysis and comparison of medical policy documents across healthcare providers. It enables efficient document parsing, knowledge retrieval, and comparative reporting, significantly reducing manual effort and accelerating decision-making.

Problem Statement

Manually analyzing, comparing, and interpreting healthcare insurance policies from different providers (like Aetna, Anthem, Cigna) is a time-consuming, error-prone process. Policy analysts, doctors, and hospital admins face significant delays when reviewing prior authorization rules, coverage criteria, and exceptions buried deep within lengthy policy documents.

Objectives

  • Automate the retrieval of the latest medical policy documents stored in Amazon S3.
  • Process, clean, and pass these documents as inputs to OpenAI's LLMs.
  • Use prompt engineering to generate accurate, comparative insights between multiple providers.
  • Deliver a seamless, user-friendly interface for policy selection and analysis via Streamlit.

Product/Project Analysis

  • Data Source: Medical policies fetched dynamically from an Amazon S3 bucket.
  • Preprocessing: Text extraction from PDFs/HTML documents.
  • Prompt Engineering: Tailored prompts based on policy type (e.g., MRI coverage) to retrieve focused, contextual outputs.
  • Comparison Engine: Summarizes key differences (e.g., coverage limits, documentation requirements, CPT codes).
  • Output: Structured, comparative tables and narrative summaries for decision-making.

Architecture

☁️ Amazon S3
(Medical Policies)
⬇️
🛠️ Document Preprocessing
⬇️
🧠 Prompt Engineering
+ OpenAI GPT Analysis
⬇️
🎛️ Streamlit Frontend
(User Interaction + Display)

Impact

  • 80% Faster: Reduces time spent on policy analysis from days to minutes.
  • Higher Accuracy: Minimizes human oversight errors by using AI-driven summarization.
  • Cost Efficiency: Cuts down manual labor costs by automating routine policy comparisons.
  • Scalability: Can easily be expanded to new insurance providers and policy formats.
  • Democratization of Expertise: Makes complex policy insights accessible even to non-legal, non-medical users.

Conclusion

This project demonstrates the potential of Generative AI in the healthcare industry by automating a critical, knowledge-heavy task: medical policy analysis. Through smart cloud integration, document processing, and LLM-based understanding, it offers a scalable, efficient, and highly impactful solution. Future enhancements will include OCR for scanned policy PDFs, fine-tuned custom LLM models for even more domain-specific insights, and enterprise-level deployment options.

Let's Connect!

I enjoy connecting with like-minded professionals passionate about technology, strategy, and impact. Feel free to reach out!

Chicago, IL
(312) 871-8022
k.teckchandani1703@gmail.com