Pfizer Accelerates Innovation with Generative AI and AWS Cloud Services

BeijingDecember 11, 2023 /PR Newswire/ -- On AWS2023 re:InventAt the global conference, Pfizer's Chief Digital and Technology OfficerLidia FonsecaInvited by AWS CEOAdam SelipskyIn the keynote speech, introduced Pfizer's ongoing generativeAIRelated work, as well as its innovative collaboration achievements with AWS over the past year and future plans.

Pfizer Chief Digital and Technology Officer Lidia Fonseca

Exploring Generative Models in the Life Sciences FieldAIThe Future: Redistributing Time and Resources to Help Patients More Quickly

In the past year, no technology has been more significant than generativeAIHas received more attention. With the powerful technical capabilities of AWS, Pfizer in17In various use cases, Pfizer drives innovation and enhances productivity. From content generation in scientific and medical fields to more scenarios like manufacturing, Pfizer leveragesAmazon BedrockAnd AWS's artificial intelligence (AI) and machine learning (ML) technology, and established multiple prototype projects within just a few weeks. Pfizer expects that some of these important use cases will save millions annually.7.5Up to10Billion-dollar cost.

For example, generativeAIWill play a key role in identifying new tumor targets. Currently, this is a manual process that requires aggregating information from different data sources. But with the help ofAI, Pfizer can identify and organize relevant data and content from more sources in less time.AIThe algorithm can not only evaluate the trends of generating potential targets but also validate them more quickly, thereby increasing the success rate.

Moreover, by using AWS services, Pfizer quickly deployed its own certified generativeAIPlatformVOX, allowing employees to innovate in a safe manner and access includingAmazon BedrockAndAmazon SageMakerThe best large language model, including the model provided on the platform. Pfizer is also actively exploring other generativeAIApplications, including drafting initial patent applications and generating medical and scientific content for human review and final confirmation, save time while accelerating the delivery of breakthrough results to patients.

The agile culture of AWS aligns closely with Pfizer's way of working, enabling the two companies to rapidly iterate on prototypes. WithAmazon BedrockWith the various large language models provided, Pfizer can select more suitable tools for specific use cases. The modular approach of AWS enables Pfizer to easily connect with different vendors, which has greatly accelerated Pfizer's overall...AIThe advancement of the strategy.

Cloud Computing Brings Unprecedented Speed and Scale: The Journey of COVID-19 Vaccine Development and the Future

Review of COVID-19 (COVID-19) During the pandemic, AWS and Pfizer actively enabled technology to play a more meaningful role. From Pfizer's announcement of its collaboration withBioNTechCo-developing the COVID-19 vaccine, to obtainingFDA(FDA) Emergency Use Authorization, took only269Days. Under normal circumstances, this process requires8To10Years. Before the outbreak, Pfizer's entire product portfolio was manufactured.2.2Hundreds of millions of doses of vaccines. And only in2022Year, COVID-19 VaccineComirnatyThe output has already soared to40Hundreds of millions of doses. This achievement requires a great deal of technical cooperation to realize.

First, AWS quickly provided additional high-performance computing power, allowing Pfizer to scale up tens of thousands of core computing resources in the cloud. Through these additionalCPU(Central Processing Unit), Pfizer can perform computation-intensive analyses to understand how to manufacture candidate vaccines. Afterwards, when Pfizer needs to submit data in less than a day forFDADuring the application process, AWS scaled up computing power on demand, allowing Pfizer to move forward quickly.

Next, Pfizer's industry-first digital operations center provides a platform for cross-factory collaboration for the team, enabling real-time monitoring of production status and problem-solving, thus achieving20%Increase in production capacity——This has become the core of Pfizer's factory operations today. For example, throughmRNAPredictive Algorithm Increases Pfizer's Vaccine Yield Per Batch210,000 doses.

Today, Pfizer and AWS are usingAITechnology generates proactive alerts to identify events that could potentially disrupt the supply chain, such as hurricanes.“Ian”Intervene in advance before it arrives to ensure the continuous supply of key drugs and vaccines.

Establishing an Enterprise Data Strategy to Drive Global Business Success

Pfizer is currently working hard to promote an unprecedented goal in the industry, which is to18To be launched within the month19Drugs and vaccines. Significant progress has been made towards this ambitious goal.——Pfizer has now successfully launched13Type.

Digitalization, Data andAIIt is crucial to achieving this goal, and what sets Pfizer apart is how it applies these technologies across the entire company. Pfizer laid the foundation for this years ago by centralizing data, creating globally scalable standards and platforms, and fostering strong digital capabilities.AITalent and building a robust and secure foundation, Pfizer for technology andAIThe vigorous development within the company has created the necessary conditions to drive more impactful innovation.

2021In that year, Pfizer launched a key initiative: to move its core operations to the cloud.ITThe proportion of infrastructure from10%Increase to80%. This requires in42Intra-week Migration1.210,000 applications and databases as well as8000AWS Server——This is one of the fastest migration projects implemented on AWS by a company of the same scale as Pfizer.

Migrating to AWS saves Pfizer more than4700Million USD, and helped the company reduce three data centers, decreasing4700Tons of carbon dioxide emissions, equivalent to1000A family's annual energy usage. These also provide Pfizer with rapid and large-scale innovation. For example, Pfizer successfully reduced the time required to create computing power from several months to just a few hours (1Hours can be extended to610,000CPU), thereby increasing the speed of data generation required for large drug submissions.75%。

And all of this is built on earlier successes:2019Pfizer and AWS Collaborate to Integrate Multimodal Data from Hundreds of Laboratory Instruments, Developing a Leading Scientific Data Cloud in the IndustryScientific Data Cloud（SDC). This end-to-end platform can simplify the processes of handling, storing, retrieving, reusing, and analyzing scientific data generated by laboratory and production equipment.SDCThe open data lake architecture built on AWS allows the platform to dynamically scale as data volume grows and analytical needs become more complex. Since its delivery, it has empowered Pfizer scientists to conduct real-time, easy, and customized searches across all historical molecular and compound data—tasks that previously took weeks or months in a fragmented environment.SDCThe time savings brought about have enabled the Pfizer team to accelerate analysis and computational research, and leverageAIAlgorithms to identify and design the most promising new compounds.