JapanFinance and insurance2025 – ongoingKyoto Financial Group
Kyoto Bank generative AI search of internal manuals
Kyoto Bank launched AI manual search after testing reported 86% search accuracy
Problem
Kyoto Bank's internal chatbot depended on manually registered FAQs. Questions without entries still required human responses, while maintaining the FAQ collection added work.
Approach
The bank added PKSHA's generative AI document-search and answer-generation function to its existing chatbot in February 2025. It began operating against approximately 1,000 rules and manuals after several months of testing. The test reported 86% search accuracy.
Results
Takeaway
Searching source documents reduced dependence on separately maintained FAQ entries. The projected annual time saving had not been reported as an achieved result.
Categories
- Industry
- Finance and insurance
- Business function
- Internal work and knowledge management
- Innovation type
- Process innovation
- AI technology
- Generative AI
- Adoption
- Vendor solution
- Evidence
- Vendor case study
Sources
Source summaries
PKSHA confirmed that document search entered operation in February 2025 and reported 86% accuracy in testing. Its expected annual saving of 8,000 hours was a forecast.
Full details are in the original sources.