Amazon Web Services and Pfizer Collaborate to Accelerate Medical Breakthroughs with Machine Learning and Analytics

BEIJING, Dec. 7, 2021 /PRNewswire/ -- AWS announced that it is collaborating with Pfizer to create cloud-based innovative solutions aimed at improving the development, manufacturing, and distribution of clinical trials for new drugs. Through the newly established "Pfizer – AWS Collaboration Team (PACT)" initiative, the two companies are exploring ways to apply AWS's capabilities in analytics, machine learning, computing, storage, security, and cloud data warehousing to Pfizer’s lab, clinical manufacturing, and clinical supply chain activities. For instance, AWS is helping Pfizer enhance its continuous clinical manufacturing processes by utilizing AWS machine learning services such as Amazon Lookout for Equipment (an AWS service that detects abnormal equipment behavior by analyzing sensor data) to build predictive maintenance capabilities. With these services, Pfizer can maximize the uptime of equipment used in clinical drug manufacturing, such as centrifuges, mixers, grinders, coaters, and air handling units. This collaboration will support Pfizer in producing new drugs more quickly and reliably while evaluating their potential benefits to patient health.

Vice President of Business Development and Industry at AWSKathrin Renz Said: "An increasing number of life sciences industry customers are looking for opportunities to massively expand their expertise and insights, while being able to securely access the right information at the right time to reduce the time and cost of drug development and clinical trials. AWS offers deep and broad cloud capabilities that help support Pfizer teams through secure and innovative research methods, enabling them to optimize drug development and clinical manufacturing processes. The past two years have shown the world that speed and agility are crucial at every step of the research, development, and clinical manufacturing cycle when lives are at stake. We are proud to collaborate with Pfizer, leveraging our extensive domain expertise to assist in developing solutions that can significantly improve the lives of patients worldwide."

Vice President of Pfizer Pharmaceutical Sciences, Global R&D and Medical Andrew McKillop Stated: "The collaboration between Pfizer and AWS aims to accelerate the process of drug discovery and development, ultimately enhancing the treatment experience for patients and bringing new therapies to market. We are working closely with machine learning and analytics experts from AWS to provide our scientists and researchers with the insights they need to drive medical breakthroughs that can improve patients' lives."

AWS is collaborating with Pfizer to develop a prototype solution for detecting abnormal data points in Pfizer's continuous clinical manufacturing platform for solid oral medications. The prototype solution utilizes Amazon SageMaker (AWS's service for quickly building, training, and deploying machine learning models in the cloud and at the edge), Amazon Lookout for Equipment, Amazon Lookout for Metrics (AWS's service for automatically detecting metric anomalies and identifying their root causes), and Amazon QuickSight (AWS's scalable, machine-learning-driven business intelligence service in the cloud). The machine learning models used in the prototype can provide early warnings, direct users to relevant warning signals, and achieve an ultra-low false positive rate. As a result, Pfizer can process data fromPortable, Continuous, Miniaturized, and Modular(PCMM) manufacturing equipment and sensor data, detecting anomalies, predicting maintenance needs, and reducing potential equipment downtime.

Pfizer Scientists will also collaborate with AWS healthcare and life sciences experts to explore how researchers from Pfizer's Pharmaceutical Sciences Small Molecule team can leverage AWS analytics and machine learning services to extract and mine information from past documents. Pfizer has accumulated a vast number of files containing valuable data generated throughout various drug development processes, covering chemical synthesis routes, formulations, analytical testing, method development, formulation composition, clinical manufacturing activities, batch records, technology transfer, and many other types of workflow-generated data. These documents contain information with powerful insights that, if effectively identified and organized, could guide Pfizer’s researchers in the development of new drugs or the repurposing of existing ones. To quickly and securely access the right information within a specified timeframe, Pfizer's Small Molecule team is working with AWS to develop a prototype system capable of automatically extracting, ingesting, and processing data from documents to assist laboratories in designing experiments. This prototype system is powered by Amazon Comprehend Medical (AWS's HIPAA-compliant natural language processing (NLP) service that accurately and rapidly extracts information from unstructured medical text) and Amazon SageMaker, and uses Amazon Cognito to provide secure user access control.