HealthEM.AI Build Cohort: 5Wk Pilot Implementation

Objective: Streamline and simplify medical and PMS data across all points of care to achieve strategic goals based on data insights and proven AI/ML models.

Key Challenges Addressed:

1. Reduce un-necessary Inpatient and emergency room utilization
2. Patients’ satisfaction on care services provided
3. Identify and Retain patients on care plan by providing right care in first few period.

How do we address your challenges:

1. Use of developed AI/ML models using amalgamated claims, EMR and SDoH data to predict member future cost and outcomes
2. Unique Retraining strategy to improve model accuracy than industry standards (2-4x higher).
3. Integrated library of modules with chronic models to support informed clinical decisions.
4. Azure Cloud-enabled PaaS components for faster and scalable deployment.

Pilot Outcome:

1. Identified the priority patients based on high cost and high utilization members
2. Developed client’s goal specific and disease cohorts
3. Ability to concentrate on high priority patients in turn reduced the patient churn and cost

Implementation Plan:

1. Week 1-2: Spent on ‘Discovery’ to understand the business and data.
2. Week 2-4: Build/configure data pipelines and execute pre-built AI/ML models. Enhance models if required.
3. Week 5: Review final output, measure accuracy, create golden record & share results.

The solution is built on native Azure components to intelligently scale using below key components:

1. Azure Data Factory: Data Ingestion and orchestration solution
2. Azure Data Lake Storage Gen2 : Scalable, cost effective storage for big data analytics
3. Azure Databricks: Fast and collaborative Apache Spark-based analytics service designed for data science and data engineering
4. Azure Key Vault: For securely storing secrets and credentials
5. Azure SQL DB : Scalable relational database service
6. Azure WebApp : Scalable, managed service which hosts the front end application