HFW Companies unifies 13 ERPs and automates its close with Microsoft Fabric

Phase 1: Unifying 13 ERPs in under 90 days

Before the platform, teams relied on manual exports and quarterly, Excel-based consolidation to produce leadership reporting: time-consuming, hard to repeat consistently, and slow to answer ad hoc questions. Estimates to unify the landscape ranged from one to three years. Working with Data Crafters, HFW delivered a Microsoft Fabric data platform that integrated 13 distinct ERP systems in under 90 days. Those systems spanned three platforms common to acquisition-driven architecture and engineering groups: Unanet (10 firms), Deltek Ajera (2), and QuickBooks (1).

To bring that data in, Data Crafters built a pipeline-first ingestion pattern in Fabric Data Factory that has run since day one: Fabric Data pipelines and Spark notebooks are the primary engines, with Dataflow Gen2 used alongside them. Unanet and Ajera connect through ODBC connectors hosted on an Azure virtual machine, and QuickBooks through a managed ingestion service. The solution follows a medallion architecture, moving data from a raw Lakehouse layer through standardization and consolidation into curated Warehouse layers, where a semantic model powers reporting in Power BI with secure access that preserves group-wide oversight while enabling appropriate visibility for local teams.

The full source-to-report pipeline runs end to end in roughly 60 minutes and refreshes daily; because the Power BI dashboard refresh is built into the same pipeline, new numbers reach users within about five minutes of a run completing. The pipeline runs on a Fabric capacity that scales from F4 to F32 only during the run window, keeping steady-state cost low while still meeting the performance bar. The result is a repeatable onboarding pattern that lets HFW absorb new firms without redesigning the platform.

As Shreena Fulani, Director of Business Intelligence, HFW Companies, adds: “For the first time ever, we have a single source of truth for our data. HFW now has the capability to analyze and gain data insights from all our partner firms in Power BI, an effort which previously took weeks to manually consolidate from various ERPs.”

Phase 2: Automating the monthly Financial Package

With the foundation in place, HFW turned to its most visible deliverable: the monthly Financial Package for the board and investors, where the standard for accuracy is absolute. The previous process consumed three full-time finance team members for two weeks every month, reconciling data by hand across 13 business units and stitching it together in Excel, with numbers sometimes reaching leadership up to two months after the period closed. The transformation started with principles, not tools: every number would have one unified definition, accuracy would be held without compromise, and finance (not technology) would own the definitions. Data Crafters and HFW extended the medallion architecture into a finance-grade platform on Fabric. The monthly Financial Package that once took three people two weeks to assemble by hand is now generated automatically from a nightly cycle that completes in under 30 minutes and can be refreshed any day. Freed from manual assembly, the team now reviews the package rather than building it and is steadily reducing the reporting lag, from roughly two months toward about four weeks and closing, on track to deliver within five to ten days of month-end. With assembly automated, finance has more time to analyze performance, investigate exceptions, and support leadership decisions.

Inside the finance-grade Fabric platform

Phase 2 didn't replace the platform, it extended it. The same Fabric Data Factory pipelines, Spark notebooks, and medallion architecture built in Phase 1 continue to move data from every ERP into Fabric; Phase 2 added a finance-focused semantic layer and user-driven data management on top. When HFW acquired another firm in April 2026, its ERP was onboarded on the same ingestion pattern and architecture, with no change to the design, because the platform built on Fabric was designed to scale. The group has continued to expand since, and each new firm now flows into the same governed model and monthly reporting without a separate integration effort. Today the finance model spans roughly 13,900 GL accounts and processes around 50,000 transactions a month.

At the top of the finance layer, Fabric Planning gives the finance team a front end to own its own logic: chart-of-account mappings apply automatically, and finance, not engineering, holds a controlled override. Any manual mapping change or adjustment is entered by finance, then proposed, reviewed, recorded, and versioned in a Fabric SQL database, with metadata owned by finance and ready for audit. From there the entries flow into the existing medallion architecture: a raw layer, a transform layer for cleanup, standardization, and unification, then a curated EDW layer feeding a finance analytics model and finance statement reports.

A validation framework reconciles the platform's trial balance against external trial balances every month, on average around 1,400 active general ledger accounts. Reconciliation now runs at zero tolerance, tightened from an initial $5,000 per account to $0, so any account-level difference is flagged with full lineage back to source, for a 100% monthly reconciliation pass rate. A Power BI semantic model drives the Financial Package, and the monthly Excel workbook is generated automatically from that same model, so the P&L, balance sheet, and cash flow all draw from one governed source.

Built with Data Crafters

HFW's finance team partnered with Data Crafters across both phases. Beyond the Fabric implementation, Data Crafters helped translate accounting standards into reconciliation rules and shaped the definitions themselves, building the platform and the finance logic together. The same standardized foundation now absorbs new companies directly into the reporting cycle: a newly acquired firm joins the centralized data model in under a month, with full reporting across the Financial Package and Revenue Segmentation report following in under two months, work that would previously have taken a year or more. As Shreena Fulani, Director of Business Intelligence, HFW Companies, notes: "When a newly acquired firm joined HFW in April 2026, we integrated it into Microsoft Fabric in less than two months and automated its data into both the Financial Package and the Revenue Segmentation report." That mix of Fabric engineering and finance-grade data modeling is what let the two teams build the platform and the accounting logic together, and turn each acquisition into a repeatable onboarding rather than a one-off project.