The task of developing, testing, and evaluating research and regulatory-grade AI systems is challenging, especially when it involves multi-modal clinical and scientific data. We will discuss our framework, processes, quality system, and technology stack to overcome these challenges and develop and deploy AI systems at scale.

Cloud computing provides vast storage and computing capabilities to life science. However, the teams managing these resources may sometimes feel detached from the actual science. Cloud teams require strategies to help optimize costs, enhance security, and modernize scientific workloads when they lack context. This talk will equip cloud leaders and professionals with tools and techniques for more effective cloud implementation, enabling both scientific innovation and fiscal prudence.

This talk is about how Vertex Pharmaceuticals modernized its image segmentation platform using AWS Serverless technologies and explains how it benefited our scientists in their research & development work. We will review how Vertex’s legacy system worked and pain points associated with it, discuss the new cloud-based system using AWS Serverless, benefits gained, and lessons learned.

Navigating various data types and multiple compute environments is crucial to modern data science in biotechnology. The session will cover how the cloud plays a pivotal role in enabling scientific research: allowing us to manage diverse datasets, leverage different computational architectures for different needs, as well as crucial tracking and integration demands for our gene writing research.

Generate:Biomedicines is pioneering the field of generative biology—using machine learning models to transform the biological drug discovery process from one of trial-and-error to one that is engineerable, predictable, and repeatable. Our cloud computing infrastructure supports our scientific teams by enabling efficient use of computational tools, highly scalable workloads, secure general accessibility, and commercial and internally developed informatics applications.

Life sciences research environments are rapidly evolving to incorporate increasingly diverse resources. Using multiple compute and storage environments efficiently requires researchers to invest valuable time becoming familiar with their unique access methods, and integrating them into research pipelines. We will present a platform that unifies access to compute and storage resources, both cloud-hosted and on premises, and demonstrate how a familiar interface provides users with a superior user experience which allows users to focus on their research rather than on technology management. We will describe and demonstrate how a novel implementation of functions-as-a-service enables even those with only rudimentary programming knowledge to easily compute at any scale, from a laptop to a supercomputer.

Join us as we delve into the evolutionary journey of cloud infrastructure at Tessera Therapeutics. We explore the early choices that kick-started our cloud journey and how those choices evolved in response to the dynamic requirements. In this talk, we emphasize how we tackled a range of challenges encountered on the path to scalability. We will cover initial infrastructure decisions, strategies for adapting to rapid growth, technology choices, and numerous experiences of overcoming challenges.

In cooperation with clinical experts, we developed a system for convenient labeling and performing analysis on videoendoscopy data. Leveraging the advantages of cloud infrastructure, ENDO-CLOUD allows for better coordination and standardization of the work of multiple international research groups. For implementing AI modules and continuous model development, the system utilizes cloud computing capabilities. Our goal is to provide cutting-edge ICT applications for clinical laryngology, both in clinics and operating rooms.