AWS re:Invent 2025: A transformative moment for healthcare and life sciences

At the 14th annual AWS re:Invent, more than sixty-thousand executives, developers, and industry leaders united in Las Vegas to discover and explore the most cutting-edge data and AI innovations. Amidst dozens of launches, hundreds of demos, and thousands of sessions, healthcare and life sciences customers took center stage. They shared real-world use cases to demonstrate how they’re leveraging Amazon Web Services (AWS) solutions to accelerate drug discovery, improve clinical workflows, and transform patient experiences.

In this blog, the AWS healthcare and life sciences team shares their top sessions, announcements, and customer spotlights.

Sessions

Healthcare and life sciences customers were featured in dozens of sessions at re:Invent. Here are some of our favorites:

• Accelerating Real-world Evidence Generation in Life Sciences
• Reclaiming Clinical Time: How Veradigm Uses AI To Transform Workloads
• Novartis’s Next-gen Data Platform Enables AI Drug Development
• UC San Diego Health Modernizes Patient Engagements with Amazon Connect
• From Lab to Market: AstraZeneca’s Enterprise-Wide AI Success Story
• Healthcare Transformed: Reimagining the Future with Generative AI
• Beyond EHR – leveraging AWS for Maximum Clinical and Operational Impact

Keynotes

In his annual keynote address, Matt Garman featured healthcare and life sciences customers as industry leading examples of innovation, including Lila Sciences, Bristol Myers Squibb, Cohere Health, and Pfizer. Matt’s emphasis on developers as “the heart of AWS” and “the freedom to invent” remains AWS’s core mission after 20 years.

Swami Sivasubramanian’s AI keynote opened with a spotlight on the Allen Institute. His talk highlighted the organization’s development of an advanced neural net model for analyzing single cell multimodal brain cell data. Swami went on to highlight several health use cases during his presentation, including training models for specific patient outcomes and developing an AI model that deeply understands molecular structures and protein interactions.

Peter DeSantis and Dave Brown reinforced the core attributes that AWS obsesses over, security, availability, performance, elasticity, cost, and agility. Even more, these foundational cloud elements are more important than ever in the AI era. Dave Brown showcased Graviton and AWS’s custom silicon innovations that deliver these attributes at scale.

Werner Vogels delivered his final keynote after 14 years, introducing the concept of the “renaissance developer” someone who is curious, thinks in systems, and communicates effectively. His message about AI and developer evolution resonated: “Will AI take my job? Maybe. Will AI make me obsolete? Absolutely not…if you evolve.” He emphasized developers must be owners: “The work is yours, not that of the tools. You build it, you own it.”

Announcements

As healthcare organizations continue their digital transformation journeys, this year’s announcements directly addressed the industry’s most persistent challenges: data privacy, clinical workflow efficiency, specialized AI development, and secure infrastructure.

Additionally, life sciences organizations continue to expand their usage of agentic AI across the value chain. New services promise to accelerate time to production while driving greater efficiencies, from target identification to personalized patient engagements.

With dozens of new services, features, and updates announced during the week, our healthcare and life sciences experts have honed in on the most impactful areas of innovation.

Data privacy and security innovations

Security remains job zero, and this year AWS launched several new services and updates to help our customers innovate while keeping their data secure and maintaining compliance. One of the most impactful launches is an expansion of AWS Clean Rooms, enabling privacy-enhancing synthetic dataset generation for AWS Clean Rooms.

With this new capability, organizations and their partners can generate privacy-enhancing synthetic datasets from their collective data to train regression and classification machine learning (ML) models. For healthcare and life sciences organizations, this update allows more tailored and fine-tuned models without exposing patient or proprietary data. The ability to generate synthetic datasets that maintain statistical validity while adding configurable privacy protection directly addresses one of healthcare’s biggest challenges: balancing data utility with privacy requirements. For example, research institutions can now collaborate on rare disease studies by generating synthetic datasets that preserve statistical patterns while maintaining HIPAA compliance. Further, drug discovery teams can train models across multiple clinical datasets without exposing patient health information (PHI.)

Additional announcements included:

• Data sovereignty: AWS AI Factories allows healthcare and life sciences organizations to harness the power of AI while adhering to industry regulatory and compliance needs. This enables local processing of sensitive patient data, improving compliance posture, and adhering to PHI requirements. For example, a healthcare organization can now deploy AWS AI infrastructure within their own hospital data centers.
• Proactively secure applications: AWS Security Agent enables proactive security development and monitoring, essential for healthcare applications handling PHI. Security Agent helps maintain HIPAA and GxP compliance through continuous security assessments. Also, it elps identify security issues before they impact patient or proprietary data.
• Threat detection: Amazon GuardDuty Extended Threat Detection provides unified security visibility across healthcare environments. This can help identify sophisticated attacks targeting patient data across diverse healthcare workloads. What is more, this new feature simplifies security monitoring for complex healthcare IT environments and help generate comprehensive security reports for HIPAA and other regulatory audits.
• Security hub: AWS Security Hub helps prioritize critical security issues and respond at scale while improving response times. This helps healthcare organizations minimize downtime disruptions with near-real-time analytics, and automate issue prioritization and compliance checks.

AI Advancements
Throughout 2025, AWS continuously invested in developing new services like Amazon Bedrock AgentCore that make it easier for all organizations to harness the power of AI. At re:Invent, we saw this trend continue with a lengthy list of announcements.

Here are a few that stood out:

• Amazon Connect launched new capabilities for patient engagement. It now features secure, real-time integration with electronic health records (EHRs), enabling self-service verification for patients and caregivers, and verifying appointments are scheduled with up-to-date, accurate information. Natural Voice Interactions with Advanced Speech Models: Powered by Nova Sonic’s advanced speech models, AI agents have natural, human-like conversations with appropriate pace, tone, and understanding across multiple languages and accents.
• AWS announced Amazon Nova 2 – the next generation of general models that deliver reasoning capabilities with industry-leading price performance.
  • Speech-to-speech: Amazon Nova 2 Sonic, a speech-to-speech model, delivers industry-leading conversational quality, pricing, and best-in-class speech understanding for developers to build voice applications. For patient interactions and accessibility, Sonic can enable more natural multilingual conversations and real-time translation for telehealth services and remote monitoring. And in the life sciences, Sonic can help researchers interrogate data sources with speech, formulate hypothesis, and capture lab work through speech rather than manual entry, saving time and increasing productivity.
  • Multimodal data: Amazon Nova 2 Lite and Nova 2 Omni provide cost-effective AI with extended reasoning capabilities. One powerful feature is the multimodal data integration – enabling integrated analysis of medical images, text reports, research literature, and patient data. With a million-token context window, Lite allows processing of entire patient histories for more informed recommendations. Additionally, for clinical trials and remote monitoring, it can process multiple data streams from wearables to home monitoring devices.
  • Develop specialized domain models: Amazon Nova Forge makes it easier to develop models for specialized domains. This is especially helpful in the healthcare and life sciences industry, from developing a drug discovery assistant that predicts molecular properties to helping health systems build custom models for radiology, pathology, or genomics that deeply embed domain expertise.
  • Automate repetitive work: Amazon Nova Act enables secure automation of repetitive workflows, such as EHR data entry, claims processing, and follow-up coordination.

• AWS EC2 Trainium3 and P6e UltraServers provide improved price-performance options for training specialized healthcare and life sciences models. This is especially beneficial when training with complex and large datasets, such medical imaging, genomics, and digital pathology.

Data management and analytics
A secure and scalable data foundation continues to be the driving force for AI success. At re:Invent, AWS announced numerous data management and analytics services. For our healthcare and life sciences customers, our top picks are vector scalability and replication.

Amazon S3 Vectors is the first cloud object store with native support to store and query vectors, delivering purpose-built, cost-optimized vector storage for AI agents, AI inference, and semantic search of your content stored in Amazon S3. For genomic analysis, researchers can efficiently store and query billions of genomic vectors while keeping costs low. In medical imaging, Vectors enable rapid similarity searches across vast imaging repositories. And in drug discovery, researchers can accelerate compound screening by quickly identifying similar molecular structures.

Amazon S3 Tables with Replication simplifies compliance with data residency requirements and disaster recovery. For multi-region compliance, healthcare companies can now easily maintain synchronized patient data across regions to meet varying regulatory requirements. For researchers, this simplifies data sharing between research sites while maintaining consistent governance.

Strategic recommendations

Based on these selected industry announcements at re:Invent, healthcare and life sciences organizational leaders should consider:

• Data strategy reassessment: Evaluate how AWS Clean Rooms and S3 Vectors can enable new collaborative research initiatives while maintaining privacy
• AI roadmap acceleration: Consider how specialized models through Nova Forge and Nova 2 could transform specific clinical or research domains
• Business process automation opportunities: Identify high-volume administrative processes that could benefit from Nova Act’s automation capabilities
• Infrastructure planning: Assess whether AWS AI Factories could address data residency and inference latency concerns that have previously limited AI adoption
• Security enhancement: Implement the new AWS Security Agent to strengthen protection of sensitive healthcare applications and data

Conclusion

AWS re:Invent 2025 represented a significant leap forward for healthcare and life sciences organizations. The focus on privacy-preserving collaboration, specialized AI model development, workflow automation, and secure infrastructure directly addresses the industry’s most pressing challenges. For healthcare organizations already on AWS, these innovations provide immediate opportunities to enhance patient care, accelerate research, and improve operational efficiency. For those still early in their cloud journey, re:Invent made a compelling case for how AWS is specifically addressing healthcare and life science’s unique requirements around privacy, security, and specialized domain expertise.