Merck inks Google Cloud agentic AI deal worth up to $1 billion

Merck will build an agentic AI platform on Google Cloud across its research and development, manufacturing, commercial operations and corporate functions in a deal valued up to $1 billion.

Under the multi-year deal, Gemini Enterprise and Google Cloud's other AI services will be deployed at Merck to enhance its digital backbone. In addition, Google Cloud will provide engineers to work alongside Merck teams to deploy AI.

For Google Cloud, the Merck win is not only big financial but serves as a model to use forward deployed engineers. In the Merck case, Google Cloud is providing engineers, but the company at Google Cloud Next also highlighted how forward deployed engineers were being utilized with partners.

• Google Cloud presses full-stack AI edge with new TPUs, Agentic Data Cloud, Gemini Enterprise Agent Platform

Merck, which operates as MSD outside the US, has more than 75,000 employees and delivered revenue of $65 billion in 2025. Merck Chief Information and Digital Officer Dave Williams said the pharmaceutical giant's Google Cloud deal is "the next phase of our AI journey" and the goal is to reimagine processes at scale.

The company has been focused on process efficiencies and reinvention for years and has been able to self-fund transformation projects through annual savings. Merck has been a long-time customer of Celonis' process intelligence and it'll be interesting to see how the company blends agentic AI with the process intelligence layer. Merck recently moved its Celonis center of excellence into the Chief AI Office.

Google Cloud CEO Thomas Kurian said the partnership with Merck highlights moves beyond process optimization and "represents a fundamental shift in how technology supports the entire pharma chain."

• Forward deployed engineers: The promise, peril in AI deployments
• AI Becomes Operational: The Board Quarterly Update

Merck's approach: AI and process

Merck operates a layered architecture that combines hyperscale cloud providers including AWS, internal AI platforms and process mining as its foundation.

In January, Merck said its AI platform is used by 80% of its workforce and it leverages generative AI for clinical documentation and workflows. The company also has foundational models for drug discovery, molecular design and toxicity prediction.

Merck added that it is looking to combine foundational models with domain knowledge across the company. Merck's AI priorities include:

• Drug discovery and unlocking patterns in datasets with AI.
• Optimizing clinical trials.
• Automating workflows across the company.
• Modernizing manufacturing.
• And streamlining education and engagement with health care providers.

In 2025, Merck outlined how it built a platform with McKinsey that uses advanced data engineering and LLMs to enhance clinical studies. The effort reduced the time to create a fully human reviewed clinical study report first draft from 180 hours to 80 hours. Errors were also cut in half.

Merck has also multiple AI initiatives that include AI chatbots for employees and field reps and text-to-SQL using Anthropic models via AWS Bedrock.

The pharma giant's addition of Google Cloud will reinforce its multi-cloud approach to AI. Merck's stack includes Celonis process mining to inform where AI is applied, a broad use of the AWS stack for infrastructure, the use of pharma-specific models and AI that's embedded into workflows.

Merck's approach is to take process understanding to orchestrate AI and ultimately automate execution. Google Cloud's deal with Merck appears to be in that latter category.

Google Cloud in its Merck announcement said it will the deal with the company includes deploying Gemini Enterprise across end-to-end R&D workflows, optimizing manufacturing, enhancing patient and customer engagement and automation in corporate functions.