Finance ops · Production

Türkiye Finans automates unstructured money transfer processing with UiPath IDP

The problem

Türkiye Finans handled money transfer instruction documents that were highly variable and unstructured, resulting in long processing times and significant manual intervention in branch operations.

First attempt

Traditional OCR engines had been tried but proved unsuccessful for processing the unstructured transfer documents.

Workflow diagram · grounded in source
1
Customer submits transfer document
trigger
“manual handling of free-format money transfer instructions from customers”
2
IDP extracts key data
ai_action
“automated the extraction of key data from the transfer forms that customers submit”
3
AI sends data to instruction system
integration
“AI tools send important information to the customer instruction system”
4
Unreliable documents held for review
human_review
“Any documents that are not reliable are kept for review”
5
Automated payment processing
output
“Today, 84% of payment instructions are processed without a human”
Reported outcome

With UiPath IDP, 84% of payment instructions are processed without human intervention, the process is significantly faster, error rates have been reduced, and the workforce can focus on more important tasks.

Reported metrics
Payment instructions processed without human intervention84%
Processing speed of payment instructionssignificantly speeding up the process
Error ratereducing error rates
Workforce focusfocus on more important tasks
Reported stack
UiPathIDP
Source
https://www.uipath.com/resources/automation-case-studies/turkiye-finans-streamlines-unstructured-money-transfer-requests-with-ai
Read source ↗

Frequently asked questions

What did this team achieve with this AI workflow?

With UiPath IDP, 84% of payment instructions are processed without human intervention, the process is significantly faster, error rates have been reduced, and the workforce can focus on more important tasks.

What tools did this team use?

UiPath, IDP.

What results were reported?

Payment instructions processed without human intervention: 84%; Processing speed of payment instructions: significantly speeding up the process; Error rate: reducing error rates; Workforce focus: focus on more important tasks (source-reported, not independently verified).

What failed first in this deployment?

Traditional OCR engines had been tried but proved unsuccessful for processing the unstructured transfer documents.

How is this finance ops AI workflow structured?

Customer submits transfer document → IDP extracts key data → AI sends data to instruction system → Unreliable documents held for review → Automated payment processing.