Novo Nordisk Scales Generative AI to 2,500+ Use Cases Across 25,000 Employees on Amazon Bedrock

Title

Novo Nordisk Scales Generative AI to 2,500+ Use Cases Across 25,000 Employees on Amazon Bedrock

Content

Novo Nordisk created a self-service platform on Amazon Web Services (AWS) for employees to innovate with generative AI and create applications for use cases in nonregulated processes across the pharmaceutical value chain. The solution uses Amazon Bedrock, a fully managed service that offers a choice of high-performing foundation models. Since implementation, more than 25,000 Novo Nordisk employees have used it to create chatbots for over 2,500 use cases to boost productivity and unlock novel applications. The company began experimenting with large language models and generative AI in 2023. Instead of trying to predict which use cases would succeed, Novo Nordisk provided an environment where employees could innovate with generative AI while maintaining security, compliance, and cost-effectiveness. The company worked with Cloud2 Oy (previously as KeyCore), an AWS Partner, to validate architectural designs. Using the generative AI platform, over 25,000 employees have created chatbots for retrieving information, drafting documents, or acting as virtual colleagues or critics. Employees receive help drafting documents where chatbots correct common mistakes and formatting issues. More than 1,000 employees use the company's off-the-shelf chatbot, which processes over 26,000 prompts in a month and moved from proof of concept to minimum viable product in one month. The largest use case was trained on 140,000 documents. Each use case costs around $10 per month to run on AWS serverless services including Amazon DynamoDB and AWS Lambda. Novo Nordisk views generative AI primarily as a catalyst for innovation and enhanced value creation, not as a means to cut costs, and is exploring use of Amazon Neptune for building knowledge graphs.