ERP Migration Case Study

Master Data Clean-up for NetSuite Migration

How we helped a company consolidate legacy ERP data when Excel failed and the Big 4 cost too much.

The Problem: Dirty Data vs. Hard Deadline

A mid-market company was migrating from Business Central (and other legacy systems) to NetSuite. They had multiple subsidiaries operating on different platforms, resulting in a massive "vendor master" and "item master" list.

The issue? The data was full of duplicates across entities, but the naming conventions weren't identical. They couldn't just migrate the data as-is, or the new NetSuite instance would be polluted on Day 1.

Why Excel Failed

  • VLOOKUP: Failed because names weren't exact matches.
  • Fuzzy Match: Produced inconsistent results (false positives) that required manual review of thousands of rows.
  • Manual Review: Estimated to take hundreds of man-hours they didn't have.

The Solution: AI-Powered Semantic Matching

We built a custom automation to ingest their data across all subsidiary ERPs and find every record that closely resembled a match using AI—not just string matching.

Deep Semantic Search

Our tool understood that "State of CA" and "California Revenue Dept" are the same entity, even if they share zero words.

Confidence Scoring

We assigned a % score to every match. 99% matches were auto-merged. 70-90% matches were flagged for quick human review.

Clean Output

The final output was a clean CSV ready for the NetSuite implementation team, with all duplicates mapped to a single "Master ID."

Real-World Matching Examples
Legacy System ALegacy System BAutomation Result
State of California Revenue DepartmentCA Dept of Rev✅ Match Identified (High Confidence)
LegalZoomLegal Zoom, Inc.✅ Match Identified (High Confidence)
Amazon Web SvcsAmazon.com⚠️ Flagged for Review (Medium Confidence)

The "Automated CPA" Difference

The "Big Guys" (Bain/PwC)
$50,000+

Quote for a similar data cleansing project.

The Automated CPA
< $10,000

Custom built. Delivered in 5 days.

"It saved them thousands of dollars and man hours and also contributed to accuracy and set the ERP migration up for success."

Timeline to Value

Day 1
Initial assessment of data complexity.
Day 3
Prototype delivered on subset of data (Free).
Day 5
Project greenlit.
Day 10
Full dataset processed and delivered for migration.

Pick one painful close task. We’ll automate it.

Assessment + prototype cost $0. If you don’t like it, walk away.