AWS re:Invent 2025 ran December 1–5 in Las Vegas and drew thousands of builders, leaders, and teams looking to sharpen their cloud strategy. The event delivered a wave of new AI tools, storage and database upgrades, and fresh compute and networking options aimed at helping organizations move faster with less friction.
Whether you’re a CTO planning next year’s roadmap or a developer trying to ship more with fewer blockers, this year’s announcements point to a cloud stack designed for higher performance, tighter governance, and better economics.
Here is our quick view of the announcements from this year’s big event!
Advancements in Artificial Intelligence and Machine Learning
Ready to push AI adoption further? AWS packed re:Invent with upgrades that expand model choice, speed up customization, and tighten security.
Amazon Nova 2 Models
AWS introduced Amazon Nova 2, a new generation of models built for tougher reasoning tasks and support for multiple data types. Teams can use them to improve automation, sharpen forecasting, and build richer AI-powered products.
Nova Sonic and Nova Forge
AWS added two new tools to support conversational and custom AI:
- Nova Sonic: A speech-to-speech system that enables natural, real-time conversations.
- Nova Forge: A platform for training models on private datasets.
AgentCore for Secure AI Automation
AWS rolled out AgentCore, a platform for managing the behavior and security of AI agents. It offers policy controls, memory tools, and evaluators to help teams deploy agents across security, operations, and customer support without losing oversight.
For CTOs: AgentCore offers a well-defined path to AI adoption with compliance built in. The use of agents is undoubtedly the next major play for enterprises and vendors alike.
On-Premises AI Infrastructure with AWS AI Factories
AI Factories bring managed AI hardware and services directly into customer data centers. Backed by Project Rainier and Trainium 2 clusters, AI Factories give enterprises cloud-scale training power locally.
Key Takeaways:
- Faster AI development with next-gen models
- Strong governance features for regulated industries
- On-prem options for latency-sensitive workloads
Innovations in Storage and Database Technologies
AWS introduced major storage and database updates to help teams cut costs, scale efficiently, and support new AI-driven workloads.
Amazon S3 Enhancements
AWS expanded S3 capabilities with new replication tools and Intelligent-Tiering for tables. S3 now provides real-time performance metrics at massive scale, including insights across billions of prefixes.
Amazon S3 Vectors
The new Amazon S3 Vectors feature supports vector indexes with up to 2 billion entries, enabling large-scale AI retrieval tasks, recommendation engines, and semantic search pipelines.
GPU-Accelerated OpenSearch Service
Amazon OpenSearch Service now supports GPU acceleration, delivering up to 10x faster vector processing at lower cost.
Database Savings Plans and Licensing Updates
RDS introduced new options for SQL Server and Oracle. Database Savings Plans and licensing optimizations allow organizations to reduce spend while maintaining availability.
For Developers: GPU-backed OpenSearch can cut query times dramatically for vector-heavy apps.
Storage Highlights:
- Lower storage spend
- Real-time performance insights
- Faster vector search
Compute, Networking, and Modernization Breakthroughs
AWS expanded its compute and networking stack to provide more control, higher throughput, and simpler modernization paths.
Lambda Managed Instances
AWS introduced Lambda Managed Instances, combining serverless automation with EC2 flexibility for stateful or persistent workloads.
EC2 Instances with Trainium 3
AWS launched new EC2 instances powered by Trainium 3. EC2 Trn3 UltraServers offer major gains for AI training and inference at lower cost.
AWS Transform
AWS released AWS Transform, an AI-driven modernization engine for Windows and mainframe apps. It accelerates code, UI, database, and testing migration workflows up to 5x.
Route 53 Global Resolver
AWS announced Route 53 Global Resolver, a global anycast DNS layer that improves resilience and latency for distributed applications.
AWS Partner Central Integration
AWS Partner Central is now embedded directly into the AWS Console, simplifying collaboration and partner management.
Compute and Networking Takeaways:
- Hybrid workloads become simpler
- Trainium 3 raises the ceiling for AI performance
- Modernization becomes less risky and more predictable
What This Means for Business Leaders
The 2025 re:Invent updates point toward a cloud environment built for faster AI adoption, smarter cost controls, and more predictable modernization.
Accelerating Digital Transformation
Leaders can use new AI models, storage enhancements, and modernization tools to build AI-driven strategies that integrate smoothly into existing infrastructure.
Balancing Innovation and Governance
AgentCore and AI Factories give organizations a way to deploy powerful AI systems while maintaining compliance and visibility.
Optimizing Costs While Scaling
Savings Plans, Intelligent-Tiering, and Trainium-powered compute options help organizations match cost with demand while improving performance.
Preparing Teams for Rapid Adoption
Training teams on Nova tools, Lambda Managed Instances, and AWS Transform equips organizations to modernize faster and deploy AI with fewer blockers.
Action Plan for 2026
For technologists, it will be important to understand what works and what doesn’t for your specific use-cases. The announcements set a good foundation for 2026 planning. Based on the updates, we think that these are some of the key steps to take to fully realize what benefits you may get with the new features and capabilities.
- Evaluate Nova 2 and Nova Forge for AI use cases.
- Pilot AgentCore to build compliant AI agents.
- Adopt S3 tiering and Savings Plans to improve cost control.
- Test Lambda Managed Instances for hybrid workloads.
- Use AWS Transform to accelerate modernization.
Leaders should evaluate where these new services fit into their roadmap and begin piloting those with the highest ROI potential. Organizations that adopt these updates early will be better positioned to scale AI, streamline operations, and strengthen compliance as the cloud landscape shifts.
Congratulations to the entire AWS engineering team on all of the exciting updates. This is a sign that they are clearly mapping a journey that their customers are keen to travel on.
