Sersight

English

Sersight

English

PUBLIC SECTOR & HEALTHCARE

Governing Enterprise AI Adoption and Optimizing Multi-Tenant M365 ROI

Governing Enterprise AI Adoption and Optimizing Multi-Tenant M365 ROI

Governing Enterprise AI Adoption and Optimizing Multi-Tenant M365 ROI

CLIENT

National Public Healthcare Authority

Illustrative UI mockup. Real client data and proprietary schemas are sanitized for confidentiality.

Enterprise AI adoption telemetry, security boundaries, and license optimization

Data & AI delivery engineered for enterprise clarity.

PROJECT STORY

The Business Challenge

Following a nationwide rollout of the Microsoft 365 ecosystem and the introduction of M365 Copilot across public healthcare institutions, executive leadership required complete auditability over software adoption and license allocation.

Key organizational and governance challenges included:

  • License Inefficiencies (Shelfware): High recurring licensing costs with no centralized visibility into inactive accounts (90+ and 180+ days of dormancy).

  • Unmeasured AI Investment: Inability to isolate paid enterprise M365 Copilot usage from standard organic web interactions to determine real AI adoption and ROI.

  • Storage & Security Vulnerabilities: Critical capacity limits on Exchange and OneDrive (>90% threshold) and unmonitored external file-sharing risks.

  • Strict Enterprise Governance: The need for fully automated, zero-trust data ingestion across high-volume telemetry logs without manual credentials.

The Technical Solution

Sersight architected an end-to-end data platform built on Microsoft Fabric (Direct Lake), PySpark, Microsoft Purview Audit Logs, and high-density SVG/DAX visual engineering:

  • Resilient Ingestion Engine (PySpark & Entra ID): Configured automated batch pipelines authenticated via Service Principals (Zero Trust). Integrated adaptive error handling to navigate Graph API rate limits (Throttling) and clean unstructured telemetry via Regex.

  • Cost-Efficient Lakehouse Storage (OneLake & Delta Lake): Designed optimized data pipelines aligned with Fabric F-SKU capacity management, enforcing automated Delta table maintenance to minimize storage footprints.

  • AI Adoption & ROI Auditing: Built specialized semantic models isolating paid Copilot application usage from free web queries, delivering clear telemetry on active vs. inactive AI seats.

  • Custom Dynamic SVG UI/UX (Power BI / DAX): Developed pixel-perfect, DAX-rendered SVG infographic cards and KPI banners, condensing complex multi-tenant telemetry into clear, executive-ready interfaces.

The Strategic Impact

The platform transformed vast telemetry logs into an executive decision-making tool for financial and security leadership:

  • Direct Cost Recovery: Immediate identification of dormant licenses, enabling strategic reallocation and significant budget optimization.

  • Accurate AI ROI Measurement: Clear visibility into Copilot engagement patterns to guide targeted change management and training initiatives.

  • Proactive Risk Mitigation: Automated alerts for critical storage thresholds and external file sharing, hardening organization-wide data governance.

MORE WORK

Explore the full Sersight case study portfolio.

Explore the full Sersight case study portfolio.

Sersight

Enterprise Data Architectures, Microsoft Fabric Lakehouses & Applied AI Solutions.

Resources

© 2025 Sersight. All rights reserved.

Sersight

Enterprise Data Architectures, Microsoft Fabric Lakehouses & Applied AI Solutions.

Resources

© 2025 Sersight. All rights reserved.

Sersight

Enterprise Data Architectures, Microsoft Fabric Lakehouses & Applied AI Solutions.

Resources

© 2025 Sersight. All rights reserved.