JapanManufacturing2023 – ongoingYanmar Holdings
Internal knowledge search with Amazon Bedrock and Kendra
Yanmar's internal-rule search saved about 1,200 minutes monthly in its engine business
Problem
Yanmar had extensive technical documentation that inexperienced employees struggled to find and reuse. Knowledge remained dependent on experienced colleagues.
Approach
Yanmar evaluated AWS models from December 2023 and began full trials in March 2024. With JMAS, it built a retrieval-augmented search system using Amazon Bedrock and Kendra. Employees asked questions in chat and received answers with source-document links.
Results
Takeaway
The reported savings concerned specific departmental workflows. Source links let employees check generated answers against internal documents.
Categories
- Size
- Large enterprise
- Industry
- Manufacturing
- Business function
- Internal work and knowledge management
- Innovation type
- Process innovation
- AI technology
- Generative AI
- Adoption
- Partner co-development
- Evidence
- Vendor case study
Sources
Source summaries
AWS described Yanmar's internal knowledge-search system and departmental use. It reported approximately 1,200 minutes saved monthly in engine-business internal-rule searches, with about 20 minutes saved per user.
Full details are in the original sources.