What challenges did EY face before AI implementation?

EY received 1.5 million payments per year from 200,000 clients, with only 30% auto-matched and cleared in SAP. The rest required 7-30 minutes to research per payment and 4-25 minutes to process manual clearing, leading to nearly an hour for close to a million payments. High manual workload led to inefficiency and increased risk of errors. Training new employees on the old process took up to two weeks. EY needed a scalable, secure, but agile process to handle complex, high-volume financial operations across global locations.

What was the impact of the AI solution?

Auto-matching and clearing improved from 30% to 80%; a further 15% matched and cleared in app. Manual intervention for payments dropped from 70% to 5%. Saved approximately 230,000 hours annually.

Source

MicrosoftExpandedProductionEvidence: Medium65/100

EY Automates Global Payment Processing for Faster and More Accurate Transactions

Use case typeCompliance automationUpdated Jun 13, 2026

Ernst & Young (EY)’s Global Finance team implemented an intelligent automation solution, ‘PowerMatch’, to improve efficiency in matching and clearing 1.5 million customer payments received annually from over 200,000 clients. Leveraging Microsoft Power Platform components such as Power Automate, AI Builder, and Dataverse, PowerMatch deeply integrates with SAP, drastically reducing manual processing. The solution features advanced data extraction, rule-based auto-matching algorithms, and seamless UI for AR teams. It improved the percentage of auto-matched and cleared payments from 30% to 80%, reducing required manual work and increasing accuracy. Employee training time also fell significantly, while global deployment is underway. The project, built in less than four months by a compact development team, demonstrates significant time savings, scalability, and business impact through low-code innovation in financial operations.

Organization: EY
Industry: Professional Services
Location: Global
Published: February 2025

Reported outcomes

Auto-matching and clearing: 30% to 80%

Other quantified impact

Manual intervention for payments: 5% to 70%Annual time savings: Approximately 230,000 hours saved annuallyErrors and rebookings: −50%

Why do we believe this?Outcome claims, sources, and evidence checks

Normalized claim

Auto-matching and clearing: 30-80% increase

learn.microsoft.comFeb 3, 2025Case studyExplicit claimMedium evidence strength

Auto-matching and clearing improved from 30% to 80%; a further 15% matched and cleared in app.

Correction recorded Jul 26, 2026 · Previously: % increase · Stored the 30% to 80% rate transition as a 50 percentage-point increase.

Normalized claim

Manual intervention for payments: 5-70% decrease

learn.microsoft.comFeb 3, 2025Case studyExplicit claimMedium evidence strength

Manual intervention for payments dropped from 70% to 5%.

Correction recorded Jul 26, 2026 · Previously: % decrease · Stored the 70% to 5% rate transition as a 65 percentage-point decrease.

Normalized claim

Annual time savings: 230,000 hours saved annually decrease

learn.microsoft.comFeb 3, 2025Case studyExplicit claimMedium evidence strength

Saved approximately 230,000 hours annually.

Correction recorded Jul 26, 2026 · Previously: 230 hours decrease · Restored the thousands separator for the annual saving.

Normalized claim

Errors and rebookings: 50% decrease

learn.microsoft.comFeb 3, 2025Case studyExplicit claimMedium evidence strength

Reduced errors and rebookings by an estimated 50%.

Correction recorded Jul 26, 2026 · Previously: 50% decrease · Relabeled the error/rebooking reduction so it is not presented as accuracy decreasing.

Why do we believe this deployment?Customer identity, provider attribution, maturity, and source checks

Customer: EY
Provider: Microsoft
Maturity: Production
Linked source: learn.microsoft.com

Deployed globally, built in under four months by a small team

Customer identity supportedSource describes one deploymentMaturity supported

Primary read

Use case focus

Showing 3 of 3

1Automated Payment Matching and Clearing for Global Accounts Receivable
2AI-Driven Entity Extraction for Payment Processing
3SAP and Power Platform Integration for High-Volume Finance Ops

EY received 1.5 million payments per year from 200,000 clients, with only 30% auto-matched and cleared in SAP.
The rest required 7-30 minutes to research per payment and 4-25 minutes to process manual clearing, leading to nearly an hour for close to a million payments.
High manual workload led to inefficiency and increased risk of errors.
Training new employees on the old process took up to two weeks.
EY needed a scalable, secure, but agile process to handle complex, high-volume financial operations across global locations.

Developed PowerMatch using Power Platform (Power Automate, AI Builder, Dataverse) to automate payment matching and clearing.
Implemented AI-driven data extraction models with a 14-step auto-matching algorithm integrated with SAP.
Used scheduled cloud flows and prebuilt SAP connectors for seamless integration.
Canvas app provided user-friendly workflows for both automatic and manual handling.
Deployed globally, built in under four months by a small team.

Technologies

Power Platform Power Automate AI Builder Dataverse SAP on Azure

Auto-matching and clearing improved from 30% to 80%; a further 15% matched and cleared in app.
Manual intervention for payments dropped from 70% to 5%.
Saved approximately 230,000 hours annually.

Architecture

The solution consists of a Power Platform canvas app (PowerMatch) using Power Automate flows to pull payment and customer data from SAP every five minutes. Payment data is stored in Dataverse. AI Builder entity extraction models process payment notifications, and a 14-step algorithm using Office Scripts applies business rules for matching. Successful matches are sent directly back to SAP for clearing; low-confidence matches are flagged in the app for manual AR team review and submission. Flows ensure bidirectional sync and update activity logs. Power Automate and custom APIs integrate with internal EY engagement and collections tools. The architecture supports high scalability, global deployment, and rapid iteration.

Sources & evidence1

Evidence: Medium65/100Evidence strength