Language
Search
-
Solutions
-
Services
- References
- Events
Agentic AI
Services
KI Agenten
LLMs
Automation Platform
Response Management
E-Invoicing
Mobile
Cloud
GenerAItor
Marketplace
Versicherungen
Banken
Gesundheitswesen
Industrie und Handel
Öffentlicher Sektor
BGV
BGV offers comprehensive insurance coverage to municipalities, businesses, and private customers. Deeply rooted in Baden, BGV stands for regional responsibility, partnership-based collaboration, and sustainable solutions.
In its central mailroom, BGV processes more than 650,000 documents annually, distinguishing between approximately 110 different business transactions. Using the Cognitive Classifier from Insiders Technologies, BGV increased the level of full automation in its input management by about 10 percentage points.
Industry
Insurance
Product
OmnIA / Cognitive Classifier
Volume
650,000 documents per year
"The new Cognitive Classifier is a significant step forward in our long-standing collaboration with Insiders. It allows us to gradually integrate AI into our existing systems and immediately achieve reliable results. This significantly reduces the effort required for clarification and has nearly doubled our productivity in post-processing."
Sören Hauser
, Team Leader, Input Management, BGV
Project Overview
What was the
initial situation?
- More than 650,000 incoming documents per year, spread across approximately 110 different business transactions
- Increasing diversity and complexity of correspondence due to varying layouts, text volumes, and embedded images
- High volumes with significant seasonal peaks, including those related to annual financial statements
- A desire for greater automated processing while maintaining stable processes and clear quality requirements
Why the decision
Insiders?
- Long-standing and successful use of smart FIX and OmnIA in input management
- Insiders Technologies’ proven industry expertise in the insurance sector
- Ability to gradually expand the existing platform with new AI components such as the Cognitive Classifier
- Confidence in the solution and the project team, even during particularly high-workload phases
How was the solution implemented?
- Expansion of the established OmnIA solution to include the Cognitive Classifier for combined text and image analysis
- Analysis of text, image, and structural features for context-sensitive recognition of business transactions
- The migration was deliberately scheduled during the annual financial reporting period—a time of increased workload—to quickly reap the benefits of the improvements
- Structured preparation and close monitoring ensured a stable transition to production
How can success
be measured?
- Increase in full automation in the incoming mail process by about 10 percentage points
- Approximately 95 out of 100 documents are correctly pre-classified and processed in the dark processing phase
- Post-processing volume reduced by just under 20 percent; productivity in the verification process nearly doubled
- Consistent turnaround times and high transparency, even during seasonal peaks
What are the next steps?
- Greater automation of additional business processes among the approximately 110
- Closer integration of related processes into the existing automation strategy
- Ongoing evaluation of new technologies with regard to benefits, cost-effectiveness, and regulatory requirements