Merck Accelerates Pharmaceutical Innovation Using AWS Generative AI and Data Analytics
Merck & Co., Inc. modernized its clinical data ecosystem to overcome siloed systems and accelerate drug development and manufacturing efficiency using AWS generative AI and analytics technologies. The company implemented a data platform spanning 300+ clinical trials and deployed AI models for medical coding, database workflows, and drug design using AWS HealthOmics and Anthropic Claude models on Amazon Bedrock. Merck also used generative AI text-to-SQL for natural language querying of healthcare data, improving analyst efficiency and accelerating R&D decisions. The manufacturing data analytics platform was rebuilt on AWS, leading to a 3x performance boost and 50% cost reduction. These innovations enabled a 70% reduction in clinical trial costs, 3x faster data processing, and overall operational cost savings.
- Organization
- Merck
- Industry
- Pharma
- Location
- United States
- Published
- April 2026
Reported outcomes
−90%
costCost savings
Strategic outcomes
Catalog median for cost savings deployments: −41% across 333 reported metrics. Compare benchmarks →
Primary read
Use case focus
Showing 3 of 4
- 1Generative AI
- 2Data Analytics
- 3Clinical Trial Optimization
- Merck faced challenges managing siloed and complex clinical and manufacturing data spanning multiple therapeutic areas.
- Extracting insights was time-consuming due to SQL complexity and disparate data sources.
- Legacy manufacturing data platforms were insufficient for increasing data demands.
- Merck collaborated with AWS and used Amazon S3, Amazon Redshift, AWS HealthOmics, and Anthropic Claude models on Amazon Bedrock to build advanced AI-powered data pipelines and generative AI solutions.
- Implemented generative AI text-to-SQL for efficient natural language data querying with over 95% accuracy.
- Built a comprehensive manufacturing data platform integrating data from hundreds of sources, automating drug design and clinical workflows with custom AI models.
- Deployed AI pipelines to improve protein design via biological foundation models.
- Achieved a 70% reduction in clinical trial costs and manual clinical work with 90% accuracy in medical coding.
- Accelerated drug development cycles with AI-powered protein design generation.
- Reduced manufacturing analytics platform costs by 50% and improved data processing speed 3x.
- Enhanced healthcare data extraction and analytics workflows, enabling faster scientific decision making.
Architecture
Merck's architecture includes a comprehensive data platform integrating clinical and manufacturing data, generative AI text-to-SQL built with Anthropic Claude models on Amazon Bedrock, AWS HealthOmics for protein design AI, and automated clinical workflows.
Sources & evidence1
AI-generated summary. Verify important details with the linked sources before relying on this case.