To reach their goal of transforming the Biopharma industry, AstraZeneca requested an architecture that could support streamlined data and analytics across the entire research and development (R&D) process.

Assessed needs. After evaluating AstraZeneca’s Data and AI program for 90 days, we identified the following needs:

• Streamlined data ingestion, processing, and advanced analytics
• The ability to analyze internal and external data sets at enterprise scale across a variety of functions like omics, drug design, imaging, and more.
• A cloud-based platform with an adaptable environment for a variety of data scientists
• The ability to search across and move data between environments
• The ability to process complex machine learning algorithms and predictive models
• Support across the entire data lifecycle: gathering and preparing data, developing and testing ML and predictive models, and deploying into production

Created a blueprint. We developed an SDF architectural blueprint that leveraged existing infrastructures and outlined data processes, architecture requirements, and project scope.

Completed a solution alignment. During this process, we aligned the architecture with the environment to ensure an integrated science foundation.

Developed and implemented a detailed plan. We identified the role of linked data in the final architecture, designed a detailed project plan with resources and timelines for a cloud-first architecture, evaluated and selected the best-fit cloud suppliers for life science AI capabilities, and documented processes and requirements.

From April - July 2019, our interdisciplinary team of consultants, architects, data scientists, semantic experts, and cloud engineers designed and implemented the Science Data Foundation (SDF).

SDF is a transparent, robust architecture that seamlessly manages data from capture to consumption and maximizes the value of data as an asset.

The SDF architecture streamlined data movement and processing at enterprise scale, ensured master and reference data consistency, and provided added context from reliable sources of truth for maximum reuse. Our hybrid cloud architecture effectively conformed and contextualized data for downstream reporting and exploratory analysis.

By applying our data governance best practice guidelines, AstraZeneca noted improved data management and process flexibility across all scientific workflows. With both on-premise and cloud-based computation capabilities, AstraZeneca accelerated time-to-market and significantly reduced costs.