How Tamr helped Amgen Build a Translational Data Platform at Scale


Identifying new therapies and finding additional applications for existing ones is a key goal for biopharmaceutical companies – as it enables company growth as well as the treatment of severe diseases. What is becoming apparent is that the role of data in this pursuit is critical and Tamr’s work with Amgen is a great point of validation to this claim.

Amgen recently embarked on a journey to build a translational data platform that enabled its scientists to analyze a multitude of studies, treatments, and specimens as well as take certain actions like requesting specimen samples for study – ultimately leading to the discovery of new therapies, new applications of existing therapies, and enabling all questions / hypotheses to be answered in a timely manner. The platform focuses on pulling critical data from disparate places (research data, biospecimen data, clinical data) and preparing it for consumption by scientists in a self-service manner who want to test a hypothesis or explore a question they have (e.g. “We know this biological pathway is important in this disease. Do patients with this marker fare better or worse with this disease? If the patients get therapy, do they respond well to it or not?”)

When constructing translational data platforms of this size and scope, traditional approaches using programmers to pull and prepare the data from many different places does not scale. It becomes very time consuming (e.g. a couple of weeks to answer one question) and it’s expensive so only the highest priority questions gets answered or worked on.

In this webinar, you’ll see how Amgen used Tamr’s human-guided machine learning to automate the incorporation of hundreds of legacy datasets. By reducing the traditional roadblocks to integrating data (cost, complexity, and time), Amgen has been able to exponentially speed up the time it takes to unify data sources across its organization and enabling all needed insight to be surfaced – ultimately leading to effective, expedited R&D decision making and rapid new hypothesis generation.