Power Automate−6 Mixed

Operator Insights

What are business and product leaders learning from rolling Power Automate out across their teams?

3 accounts: 3 reported by others

Reported by others3

Saeed ContractorHead of Technology, Intelligent Automation CoEUber
Organization-wide. Intelligent Automation CoE. Migration from an on-premises RPA platform, finished at the end of 2024.
Microsoft customer story, published Dec 3, 2024Read the case study (opens Microsoft customer story in a new tab)
Observation
Uber left its on-premises RPA platform for Power Automate and reports multimillion-dollar savings a year.
Takeaway
When a platform migration forces a rebuild, move what can run as cloud flows off desktop bots, and keep desktop flows for the screens that have no API.
What happened
When Uber left its on-premises datacenters, it reassessed its RPA platform and chose Power Automate, which it already reached through Microsoft 365. With Accenture, it migrated more than 50 processes, including safety reporting, tax categorization of menu items, vendor applications, and invoice processing. About 70% of the old bots became desktop flows on Google Cloud virtual machines and 30% became cloud flows. Uber reports over 300,000 hours a year of manual work removed and more than US$9 million saved a year. Compute costs fell by more than half for the flows moved to the cloud.
- Vendor case study
- Savings figures reported by Uber and Microsoft
- Single program
Deployment details
Organization-wide. Intelligent Automation CoE. Migration from an on-premises RPA platform, finished at the end of 2024.
- Vendor case study
- Savings figures reported by Uber and Microsoft
- Single program
Team
Uber Intelligent Automation CoE, with Accenture
Duration
Migration completed at the end of 2024
Workload
More than 50 processes: safety reporting, tax categorization, vendor applications, invoice processing
Product version
Desktop flows on Google Cloud VMs, cloud flows, AI Builder invoice model
Measurement
Hours of manual work removed and annual savings, reported by Uber

CoreLogic Asia PacificMar 10, 2025Microsoft customer storyCoreLogic rebuilt its automations on Power Automate in four months to beat a licensing renewal.
Michael O'BrienSenior Manager, Robotics AutomationCoreLogic Asia Pacific
Automation team. Property data company in Australia and New Zealand. About 5,000 employees.
Microsoft customer story, published Mar 10, 2025Read the case study (opens Microsoft customer story in a new tab)
Observation
CoreLogic rebuilt its automations on Power Automate in four months to beat a licensing renewal.
Takeaway
Before migrating, retire unused flows and merge duplicates. Build shared functions so each flow does not need its own implementation.
What happened
A licensing change on its previous RPA platform in 2023 gave CoreLogic six months to move before the next license cycle. The team merged and retired flows first, then built shared functions, such as one login routine, that every flow reuses. It moved 16 automations in four months, wrapped every desktop flow in a cloud flow, and adopted Microsoft's work queues mid-project and skipped building its own. CoreLogic reports a fivefold cut in automation costs and more than 50,000 hours of capacity returned to staff.
Vendor case study
Cost and capacity figures reported by CoreLogic
Single program
Deployment details
Automation team. Property data company in Australia and New Zealand. About 5,000 employees.
Vendor case study
Cost and capacity figures reported by CoreLogic
Single program
Team
CoreLogic Asia Pacific automation team
Duration
Four-month migration
Workload
16 migrated automations, then 87% of a corporate team's administrative work
Product version
Desktop flows run concurrently and embedded in cloud flows, work queues
Measurement
Platform cost against the previous RPA platform; hours returned to staff
EYNov 21, 2025Microsoft customer storyEY's global finance team raised automatic payment matching from 30% to 80% with flows and AI Builder over SAP.
Marie SandersGlobal Process Owner, Accounts ReceivableEY
Global finance. About 1.5 million client payments a year. Power Apps and Power Automate over SAP.
Microsoft customer story, published Nov 21, 2025Read the case study (opens Microsoft customer story in a new tab)
Observation
EY's global finance team raised automatic payment matching from 30% to 80% with flows and AI Builder over SAP.
Takeaway
Connect flows to the system of record through its connector, measure the share of items that clear without a person, and roll out region by region.
What happened
EY's finance team built PowerMatch, a canvas app with Power Automate flows that use AI Builder models to read payments and remittance advices and match them to invoices. Automatic matching and clearing rose from 30% to 80% of payments, and only 5% still need manual processing. Rolled out to about half of EY, it saves 120,000 hours a year, with about 230,000 expected at full rollout. A related app, PowerPost, posts journals to SAP through a cloud flow and the SAP connector and cut general ledger lead times by 95%.
Vendor case study
Matching rates and hours reported by EY
Single program
Deployment details
Global finance. About 1.5 million client payments a year. Power Apps and Power Automate over SAP.
Vendor case study
Matching rates and hours reported by EY
Single program
Team
EY Global Finance
Duration
PowerMatch built in under four months
Workload
Payment matching and clearing; general ledger postings to SAP
Product version
Power Automate cloud flows, AI Builder, SAP connector, Power Apps, Copilot Studio
Measurement
Share of payments matched automatically; hours a year

Reported by others3

Uber left its on-premises RPA platform for Power Automate and reports multimillion-dollar savings a year.

CoreLogic rebuilt its automations on Power Automate in four months to beat a licensing renewal.

EY's global finance team raised automatic payment matching from 30% to 80% with flows and AI Builder over SAP.