Claims processing · Production

Five Generative AI Use Cases for Enterprise IT Leaders — Appian

The problem

Enterprises broadly face challenges bridging the gap between AI anticipation and implementation, while individual organizations struggled with labor-intensive, error-prone manual processes across claims handling, customer service, and document processing.

First attempt

Leroy Merlin's fragmented oversight and data silos between systems caused prolonged approval processes, frequent order cancellations, a decline in customer satisfaction, and financial losses from inaccurate data. Aviva France's agents dealt with disjointed systems that slowed claims handling.

Workflow diagram · grounded in source
1
Claims FNOL intake
trigger
“implemented a claims portal and First Notice of Loss (FNOL) intake process in less than 12 weeks”
2
Real-time claims AI analysis
ai_action
“Gen AI further enhances the claims process by analyzing claims data in real time, identifying potential fraud, and reducing the risk of claims leakage”
3
Historical pattern identification
ai_action
“Gen AI can analyze historical claims data and identify patterns to expedite claims processing and decision-making”
4
Employee chatbot knowledge query
trigger
“allows employees to mine proprietary documents in a centralized system where information is viewable by more than 10,000 stakeholders”
5
Job code AI recommendation
ai_action
“AI-powered application takes the unstructured text of job descriptions and instantly recommends the top three codes for that job”
6
RPA refund transaction automation
integration
“leveraging robotic process automation (RPA), Leroy Merlin automated refund payment transactions, seamlessly capturing data from various payment portals”
7
AI document processing
ai_action
“AI-powered document processing streamlined up to 90% of their manual processes, boosting customer satisfaction and propelling business growth”
Reported outcome

AI-powered workflows delivered measurable results across multiple organizations: Aviva France saw same-day claims processing surge from 1% to 25% and settlements increase by 530%; Leroy Merlin streamlined up to 90% of manual processes; Texas DPS made information viewable by more than 10,000 stakeholders; and Global Excel implemented a claims portal and FNOL intake process in less than 12 weeks.

Reported metrics
same-day claims processing rate (Aviva France)from 1% to 25%
claim settlements (Aviva France)530%
manual processes streamlined (Leroy Merlin)up to 90%
stakeholders with system access (Texas DPS)more than 10,000
Show all 7 reported metrics
same-day claims processing rate (Aviva France)from 1% to 25%
claim settlements (Aviva France)530%
manual processes streamlined (Leroy Merlin)up to 90%
stakeholders with system access (Texas DPS)more than 10,000
claims portal implementation time (Global Excel)less than 12 weeks
annual customer inquiries handled (Aviva France)80,000
application processing time (government agency)significantly reduced the time it takes to process applications
Reported stack
AppianAppian Connected ClaimsRPAlarge language models
Source
https://appian.com/blog/acp/process-automation/generative-ai-use-cases
Read source ↗

Frequently asked questions

What did this team achieve with this AI workflow?

AI-powered workflows delivered measurable results across multiple organizations: Aviva France saw same-day claims processing surge from 1% to 25% and settlements increase by 530%; Leroy Merlin streamlined up to 90% of…

What tools did this team use?

Appian, Appian Connected Claims, RPA, large language models.

What results were reported?

same-day claims processing rate (Aviva France): from 1% to 25%; claim settlements (Aviva France): 530%; manual processes streamlined (Leroy Merlin): up to 90%; stakeholders with system access (Texas DPS): more than 10,000 (source-reported, not independently verified).

What failed first in this deployment?

Leroy Merlin's fragmented oversight and data silos between systems caused prolonged approval processes, frequent order cancellations, a decline in customer satisfaction, and financial losses from inaccurate data.

How is this claims processing AI workflow structured?

Claims FNOL intake → Real-time claims AI analysis → Historical pattern identification → Employee chatbot knowledge query → Job code AI recommendation → RPA refund transaction automation → AI document processing.

Related claims processing cases