A Semiconductor Manufacturer · Smart Material Selection AI Assistant “Yan Shifu”

R&D knowledge-base agent integrating technical documentation across six major product lines

Project Background

R&D knowledge-base agent: integrates technical documentation, design specifications, and historical project materials across six major product lines to enable semantic search and solution recommendations, improving new-project R&D efficiency by 30%.

Solution

  • After the first four scenarios went live, total annual operating-cost savings reached nearly RMB 10 million.
  • The R&D knowledge base is used by 90% of engineers, and the duplicate-design rate dropped by 40%.
  • Selection accuracy increased from 78% to 96%. Average selection cycle shortened from 3 days to 2 hours. Redesign iterations caused by selection errors were reduced by 70%, accelerating time to market.

Customer Testimonial

AI BootCamp truly helped us bridge the 'last mile' from strategy to execution, bringing tangible efficiency gains.
Architecture Diagram