Architecting a High-Performance In-House PACS & RIS Platform

Executive Summary

I led the end-to-end design and implementation of a proprietary PACS (Picture Archiving and Communication System) and RIS (Radiology Information System). This strategic initiative was designed to dismantle reliance on expensive foreign vendor solutions and build a specialized, high-velocity healthcare ecosystem.

By leveraging Next.js (SSR), Go, and Kubernetes, we built a platform that not only matched commercial features but exceeded them in integration flexibility. The project was awarded the Kalbe Innovation Convention 2024 for its significant contribution to digital transformation and cost optimization.

Strategic Impact

The platform transformed radiology operations from a cost-center into a streamlined, internally-owned asset.

• Financial Excellence: Achieved a 60% reduction in licensing and operational costs by migrating from legacy foreign vendors to an in-house stack.
• Operational Velocity: Reduced the time-to-integration for new hospital deployments by standardizing the interoperability layer.
• Regulatory Leadership: Fully compliant with Indonesian healthcare regulations (TKDN, CPAKB, and AKD certification), facilitating smooth local adoption.
• Engineering Ownership: Eliminated “Black Box” vendor dependencies, allowing for rapid feature deployment and custom clinical workflows.
• Award-Winning Innovation: Recognized as the top internal innovation at the Kalbe Innovation Convention 2024.

System Architecture

The architecture was designed as a cloud-native, edge-ready platform. While development and CI/CD occur in Azure, the production environment is optimized for on-premise Kubernetes clusters to meet strict data residency requirements.

It utilizes our core interoperability engine as its backbone: Universal Healthcare Integration Middleware

flowchart TD
    subgraph "Clinical Sources"
        Modality[Imaging Modalities<br/>CT, MRI, X-Ray]
    end

    subgraph "Radiology Core (On-Prem)"
        DICOM_Gate[DICOM Interface]
        PACS_Engine[MIMPS / PACS Storage]
        RIS_Engine[RIS Management]
        WebApp[Next.js SSR Frontend]
    end

    subgraph "Integration Layer"
        Middleware[Integration Middleware]
        HIS[Hospital Information Systems]
    end

    Modality -->|DICOM C-STORE| DICOM_Gate
    DICOM_Gate --> PACS_Engine
    WebApp -->|Diagnostic Viewing| PACS_Engine
    WebApp -->|Workflow UI| RIS_Engine
    RIS_Engine <-->|Study Metadata| PACS_Engine
    RIS_Engine <-->|HL7 FHIR| Middleware
    Middleware <--> HIS

Core Engineering Pillars

1. PACS (MIMPS) Engine: A robust DICOM-compliant storage and retrieval system that handles massive imaging studies with high availability.
2. RIS (Workflow Management): Orchestrates patient scheduling, exam tracking, and radiologist reporting.
3. Modern Frontend (Next.js SSR): A high-performance web interface designed for radiologists. Using SSR ensures that Sensitive Patient Data (PHI) is processed on the server, minimizing the browser footprint and enhancing security.
4. Hybrid Infrastructure: Orchestrated with GitHub Actions, moving from Azure development environments to On-Premise Kubernetes production environments.

The Challenge: Breaking the “Black Box”

The primary challenge was the High Barrier to Entry in healthcare tech. Legacy vendors often provide “Black Box” systems that are:

• Slow to integrate with local EMRs.
• Prohibitively expensive to scale.
• Difficult to customize for local regulatory needs (SATUSEHAT integration, local reporting formats).

Technical Solutions & Decisions

1. Hybrid Interoperability

Decision: Instead of building point-to-point links, I implemented a Middleware-First approach. Result: The RIS/PACS can now “talk” to almost any Hospital Information System (HIS) using a standardized FHIR/HL7 bridge, making deployments modular and predictable.

2. Security via Server-Side Rendering

Decision: Chose Next.js SSR specifically for clinical compliance. Result: By performing data fetching on the server, we eliminated direct API exposure from the client-side. This architecture ensures that even in complex hospital networks, the attack surface for PHI (Protected Health Information) is kept to a minimum.

3. Kubernetes at the Edge

Decision: Automated resource management using Kubernetes for on-premise hospital servers. Result: This allows us to provide “Cloud-like” reliability—including self-healing and easy updates—even when the system is running on a physical server inside a hospital’s basement.

Final Results

The project marked a turning point in the company’s digital strategy. We successfully delivered:

• 70% Cost Efficiency: Saved significantly on foreign currency outflows.
• Zero-Vendor Dependency: We own every line of code, from the DICOM handler to the UI.
• Scalability: The system is active in production, serving thousands of studies with high uptime.

Winner: Kalbe Innovation Convention 2024
“A technical masterpiece that proves internal engineering can outperform global vendors while delivering massive business value.”

Technology Stack

• Frontend: Next.js, Tailwind CSS, Svelte (Internal Tools)
• Backend: Go, Node.js, Integration Middleware
• Interoperability: HL7 v2, FHIR, DICOM
• Infrastructure: Kubernetes, Azure, Docker
• DevOps: GitHub Actions, ArgoCD
• Regulatory: TKDN, CPAKB, AKD Certified