Bayer China has built Dr. Bei, a generative AI training system for its pharmaceutical medical representatives, with Amazon Web Services, according to an AWS customer case study. Traditional training for medical representatives was time-consuming, costly and not customised enough, and Bayer China required any generative AI innovation to stay within a secure, manageable compliance framework. Its Digital Transformation and IT team evaluated providers and chose AWS.
Dr. Bei is a conversational bot built on the Claude large language model through Amazon Bedrock. It plays the role of a doctor in natural-language voice conversations, letting representatives practise visits in 14 different scenarios. AWS Lambda provides serverless, elastic scaling to support 450 to 1,000 concurrent users, and Amazon API Gateway manages traffic during peaks. Data access is managed through AWS Landing Zone, and Amazon Bedrock provides secure access to the models; AWS cites its 143 global compliance certifications.
The system took 72 days from proof of concept to official launch. It covered 450 users at launch and more than 1,000 within six months, and it has saved 1,000 medical representatives and trainers 6,000 to 7,000 hours of training time compared with one-on-one role-play coaching, while raising practice and coaching efficiency by more than 50%. "To accelerate our business innovation, we choose AWS for its seamless local-to-global integration and responsive support," said Patrick Moeller, CIO of Bayer Pharma.
After its results in China, Bayer has begun a global rollout of the system, with deployment under way in Europe and being advanced in several other regions across Asia and beyond.
Details
- City
- Beijing
- Organization
- Bayer
- Continent
- Asia
- Country
- China
- Category
- Pharmaceuticals