Hr ops · Production

Covestro automates sick leave certificate processing with UiPath RPA and AI, achieving 50% straight-through processing at 95% accuracy

The problem

Covestro's HR team manually processed hundreds of sick leave certificates per week to comply with German government regulations, spending up to 7 minutes per document navigating SAP to locate each employee and input data. Poor document quality made extraction difficult, and the high-volume repetitive work displaced higher-value HR activities.

Workflow diagram · grounded in source
1
Sick leave document received
trigger
“The documents are coming as scanned PDFs or pictures”
2
OCR digitization
ai_action
“at first, OCR is used to make the text machine-readable”
3
Document classification
ai_action
“The next step is classification, which helps determine the document type – whether a certain document is a sick leave certificate or not”
4
ML data extraction
ai_action
“For data extraction, a machine learning (ML) model has been created in UiPath AI Center. It is trained to process sick leave certificates in German. More than 200 documents were used to train the model to understand such fields as person…”
5
Low-confidence human review
human_review
“Whenever the extraction (or classification) confidence score is low, the output is sent for human validation via UiPath Action Center. The HR specialist can help confirm or correct the extracted data or send it for rejection in the case …”
6
SAP absence form submission
integration
“the robot logins to SAP, finds the employee profile, and fills in all needed information to submit an absence form”
7
Execution report generated
output
“Finally, the robot generates and sends an execution report”
Reported outcome

The automation achieves 50% straight-through processing with about 95% accuracy, reducing exception validation to 1–2 minutes per document from up to 7 minutes of manual work.
The HR team has reclaimed capacity to work on larger projects previously deprioritized due to daily compliance processes.

Reported metrics
Straight-through processing rate50%
Extraction accuracyabout 95%
Manual processing time per request (before automation)up to 7 minutes
Validation time per exception (after automation)1-2 minutes
Show all 5 reported metrics
straight-through processing rate50%
extraction accuracyabout 95%
manual processing time per request (before automation)up to 7 minutes
validation time per exception (after automation)1-2 minutes
HR team time savingsconsiderable time savings
Reported stack
UiPath Automation PlatformUiPath Document UnderstandingOCRUiPath AI CenterUiPath Action CenterSAP
Source
https://www.uipath.com/resources/automation-case-studies/covestro-leverages-ai-to-streamline-repetitive-hr-processes
Read source ↗

Frequently asked questions

What did this team achieve with this AI workflow?

The automation achieves 50% straight-through processing with about 95% accuracy, reducing exception validation to 1–2 minutes per document from up to 7 minutes of manual work.

What tools did this team use?

UiPath Automation Platform, UiPath Document Understanding, OCR, UiPath AI Center, UiPath Action Center, SAP.

What results were reported?

Straight-through processing rate: 50%; Extraction accuracy: about 95%; Manual processing time per request (before automation): up to 7 minutes; Validation time per exception (after automation): 1-2 minutes (source-reported, not independently verified).

How is this hr ops AI workflow structured?

Sick leave document received → OCR digitization → Document classification → ML data extraction → Low-confidence human review → SAP absence form submission → Execution report generated.

Related hr ops cases