The Decision
Growth through acquisitions needed a data platform built to scale with every deal. Microsoft Fabric was the answer.
The client is a leading U.S.-based home and building services company delivering HVAC, plumbing, electrical and energy services. Its growth strategy is built on acquisitions, backed by clear targets for EBITDA, customer satisfaction (CSAT) and operational excellence.
Each deal adds new field service, finance, payroll and marketing systems. Leadership wanted one place to see every brand side by side, and a platform that could take on the next acquisition without starting over. Microsoft Fabric was chosen as the foundation because it covers ingestion, storage, transformation, reporting and governance in one unified SaaS platform, and the company partnered with Apexon to build it.
The Problem, Reframed
The data existed in every brand. A unified, governed layer to bring it together did not.
Field service data sat in ServiceTitan, finance in Sage Intacct and payroll in ADP, alongside several marketing and operational tools. Each system used its own structure and definitions, so keeping KPIs consistent across brands was slow and difficult, and leaders could not easily compare performance across brands and business units.
Two needs shaped the design. First, every new acquisition had to join the platform through a repeatable pattern rather than a one-off project. Second, as more teams relied on the data, changes to pipelines and reports had to reach production in a controlled, consistent way instead of through manual releases.
The Work
Apexon built the entire platform on Microsoft Fabric, from ingestion to reporting to release.
Apexon designed a Fabric-native platform with OneLake as the single storage foundation for all brands. Fivetran and Fabric Data Factory pipelines ingest data from ServiceTitan, Sage Intacct, ADP and other sources into Fabric Lakehouse, organized as a landing-to-gold Medallion architecture on Delta and Parquet. Fabric Notebooks apply standardized business logic once, and semantic models turn the gold layer into consistent, business-ready KPIs.
Governance is part of the design, not an add-on. Role-based access (RBAC), row-level and object-level security (RLS/OLS), data cataloging and real-time Fabric monitoring control who sees what and keep the platform healthy. Power BI sits on the same governed layer, giving leadership and brand teams dashboards, structured reports and self-service analysis.
The latest milestone is a fully automated CI/CD pipeline, now live in production. Fabric pipelines, notebooks, lakehouse objects and Power BI reports move from Development to Test to Production through one approved, automated path. Every release is reviewed, tracked and deployed the same way.
What Changed
Microsoft Fabric is now the single source of truth for every brand, and every release reaches production automatically.
Operational, financial and marketing data from core entities and acquired brands now lives in one governed OneLake layer. Leaders track near real-time KPIs from Power BI and compare brands on the same definitions, without reconciling numbers by hand. Brand teams answer their own questions through self-service analysis.
The platform is built for the next deal. New brands follow the same Fabric pattern: connect the sources, land the data, reuse the notebooks and models, and publish to Power BI. With CI/CD in production, platform changes no longer depend on manual steps or a few key people, so the data team can deliver faster with less risk.
The result is a Microsoft Fabric implementation that is governed by design, automated end to end and ready to grow with every acquisition.
Outcomes
One Fabric platform
Ingestion, lakehouse, modeling, reporting and governance on Microsoft Fabric.
Core systems: 3+ into 1 ↓
ServiceTitan, Sage Intacct, ADP and marketing data unified in OneLake.
Automated releases: 100% ↑
CI/CD live in production across Development, Test and Production.
Speed to insight ↑
Near real-time KPIs on consistent definitions for every brand.
Acquisition-ready ↑
Repeatable Fabric patterns for onboarding each new brand.
Supporting the business story
OneLake as the single foundation. One copy of governed data for every brand, shared by every Fabric workload.
Medallion Lakehouse with reusable logic. Landing-to-gold layers, Fabric Notebooks and semantic models built once and reused.
Automated Fabric CI/CD. Pipelines, notebooks, lakehouse objects and reports promoted to production through one approved path.
Why Microsoft Fabric, and why built this way
An acquisition-led company needs a platform that stays simple as it grows. Each choice below kept the design on one platform, one pattern and one release path.
The struck-through label shows the route the company moved away from; the label below it shows the Fabric approach that replaced it.
SEPARATE TOOLS PER BRAND
Why run every brand on one Fabric platform?
Microsoft Fabric brings ingestion, storage, transformation, reporting and governance together in one SaaS platform, with OneLake holding a single governed copy of the data. Instead of stitching together separate tools for each brand, every workload reads from the same layer, so KPIs match and brands can be compared side by side.
ONE-OFF ONBOARDING
How does a new acquisition join the platform?
Each new brand follows the same steps: Fivetran and Fabric Data Factory pipelines connect its sources, data lands in the Medallion Lakehouse, and existing notebooks and semantic models are reused. Onboarding becomes a known pattern rather than a new project, which keeps effort low as the company grows.
MANUAL DEPLOYMENTS
How do changes reach production safely?
The CI/CD pipeline promotes Fabric pipelines, notebooks, lakehouse objects and Power BI reports from Development to Test to Production through one approved path. Releases no longer depend on manual steps or individual knowledge, and every change is reviewed, tracked and deployed the same way.
The Broader Implication
On Microsoft Fabric, each new acquisition adds to the picture leaders already trust.
For companies that grow by acquisition, the value of a data platform is measured at the next deal. If every brand arrives with its own systems and nothing connects them, the business becomes harder to read with each purchase.
This implementation shows how Microsoft Fabric changes that. One governed OneLake layer, repeatable onboarding patterns and an automated CI/CD path turn integration into routine work. The platform scales with the business, and leaders keep one consistent view of performance across every brand.
