Building the AI-Native Healthcare Integration Platform That Unified ArcheHealth's Fragmented Operations Across ERP, SCM, and HRM

Foundation and Data Architecture

The first problem was scope. DeVero's codebase was large, test coverage was low, and every customer ran their own application stack. A big-bang rewrite would have broken the business. Ideas2IT designed a phased migration that kept the platform live throughout.The first problem was ingestion at scale across disconnected systems. Every health system runs a different ERP, a different SCM, a different HRM, and none of them produce data in the same format. Ideas2IT built an event-driven pipeline that handles all of it without breaking compliance.

‍

1. Multi-tenant architecture with isolated schemas per customer, ensuring data security and scalability as new health systems onboard
2. Event-driven ingestion pipeline routing customer data through AWS Lambda into PostgreSQL via Databricks, with S3 as the core storage layer
3. AWS KMS encryption, IAM access controls, and a full data governance and catalogue layer maintaining HIPAA compliance across all tenants

This phase produced

• Multi-tenant HIPAA-compliant data architecture
• Event-driven ingestion pipeline
• AWS Lambda and S3 data layer
• Data governance and catalogue
• Role-based access controls and KMS encryption

Process Mining and Supply Chain Intelligence

With the data architecture in place, Ideas2IT integrated BusinessOptix for process mining and built the AI and ML layer that turns operational data into supply chain and procurement intelligence health system operators can actually act on.

1. BusinessOptix integration performing process mining on data mart outputs to surface procurement inefficiencies and supply chain bottlenecks across the full procure-to-pay workflow
2. Databricks ML models running recommendation engines and causal analysis on process-mined data to identify cost reduction and resource optimization opportunities
3. Near real-time DOMO dashboard surfacing as-is and to-be process views for business teams, with no technical expertise required to operate

This phase produced

• BusinessOptix process mining integration
• Databricks ML model layer
• Recommendation engine and causal analysis
• Near real-time DOMO BI dashboard
• As-is and to-be process visualization
• Procure-to-pay workflow intelligence

Modular Product and Go-to-Market Platform

The final layer was productization. Ideas2IT built the Arche Health App as a modular Master Data Model with a data warehouse at its core and wrapped it in a streamlined approval workflow and business rules engine that health system operators could run without engineering support.

1. Modular Master Data Model connecting ERP, SCM, and HRM data sources into a single governed data warehouse
2. Approval workflow and business rules engine with REST API endpoints for seamless integration with external health system platforms
3. OpenSearch-powered analytics layer with Kibana and Grafana monitoring, deployed on AWS EKS and ECS Fargate with full Terraform infrastructure-as-code

This phase produced

• Arche Health App with modular MDM
• Data warehouse and data mart layer
• Approval workflow and business rules engine
• REST API integration endpoints
• OpenSearch analytics
• Kibana and Grafana observability
• Terraform infrastructure-as-code