.png)
AWS Re:invent 2025 is the major annual event organized by Amazon Web Services, where the company showcases its latest technological innovations, including artificial intelligence, cloud, security, security, security, security, security, security, infrastructure, infrastructure, and systems modernization.
During this 2025 edition, Amazon Web Services unveiled a series of innovations focused on generative AI, the modernization of IT systems and the optimization of infrastructures.
The objective is clear: to enable organizations to accelerate their digital transformation while controlling their costs and strengthening their competitiveness.
The event put forward an ambitious vision, where multimodal models, autonomous agents, next-generation infrastructures and migration tools rely on seamless integration and great flexibility. These announcements confirm that AWS not only wants to keep up with the evolution of AI, but also to become a strategic innovation driver for global businesses.
The star of this edition was the Amazon Nova 2 family, an extended suite of multi-modal AI models that can understand and generate text, images, video, and audio with a much higher level of precision than previous generations. The Nova 2 Lite, Pro, Sonic and Omni versions meet various uses: automating simple tasks, generating creative content, complex analyses or assisted decision-making.
Businesses now benefit from a tool capable of processing huge volumes of data thanks to a context window of up to one million tokens. This ability opens the door to new scenarios: comprehensive audit of documents, generation of strategic reports, detection of visual elements in videos, analysis of conversations or even simulation of events.
AWS is strengthening its Bedrock platform, which is becoming one of the most comprehensive environments for AI models. Now, businesses can access an expanded set of open-weight models from a variety of partners. This openness offers more freedom to compare performance, adjust models and customize uses according to business needs.
Bedrock also integrates new tools focused on multimodal generation, automated information extraction, intelligent search and advanced management of AI agents. Thanks to these developments, organizations can create, supervise, and deploy their own agents that can perform complex tasks, such as workflow management, document writing, presentation construction, or marketing data analysis.
AWS also introduced Trainium 3, a chip designed to provide faster and more effective training for AI models. Integrated into the new UltraServers, it makes it possible to reach a higher level of performance while reducing energy consumption. With this technology, businesses can train custom models, even large ones, in shorter time frames and with reduced costs.
This advance paves the way for more customizable AI, allowing structures to adapt their models to specific jobs: legal analysis, financial forecasting, marketing segmentation, industrial automation or intelligent customer support.
Among the most strategic announcements, AWS unveiled the AI Factories, a solution for installing AI infrastructure directly in business data centers. This system is aimed at organizations subject to strict sovereignty, confidentiality or regulatory compliance constraints, such as banks, insurance companies or public institutions.
AI Factories combines the power of AWS technologies with the security of a local environment. They facilitate the deployment of advanced AI tools, while ensuring total control over sensitive data. This approach meets a growing demand from businesses seeking to adopt AI while complying with legal and internal requirements.
AWS also highlighted its new tools for automatically modernizingLegacy Systems, grouped under the name AWS Transform. Thanks to AI, these tools make it possible to convert outdated Windows applications, mainframe environments, or SQL databases more quickly to modern, cloud-compatible architectures.
AI plays a central role in analyzing code, detecting dependencies, proposing optimizations, or generating code that is rewritten in newer languages. Businesses can thus save several years of work and drastically reduce the risks associated with technical migrations.
The innovations presented at Re:invent 2025 show that AI is no longer reserved for technical experts or large structures. AWS wants to provide “ready-to-use” tools for SMEs, e-commerce, marketing agencies, and creative organizations. The possibilities are numerous: automating repetitive tasks, generating visual and textual content, generating visual and textual content, predictive analytics, improving customer service, optimizing logistics, optimizing logistics, optimizing logistics, optimizing logistics or creating digital prototypes.
By making AI more accessible, more powerful, and more secure, AWS positions its services as a solid foundation for businesses that want to accelerate their transformation without increasing investments.
