AWS re:Invent 2025 wrapped with three powerhouse closing sessions on December 5 and one big launch of Experience-Based Acceleration (EBA) partners.
Peter DeSantis on AI-driven cloud infrastructure, Deepak Singh on agentic developer tools, and Dr. Werner Vogels' final keynote unveiling the "Renaissance Developer." These talks reinforced AWS's vision: AI accelerates innovation without replacing human ingenuity, demanding evolution in cloud scalability and developer skills.
New EBA certified partners
AWS built the EBA Partner Certification Program through 2025. And they announced it during day 4. From the thousands of AWS partners around the world, AWS has selected 70 partners worldwide who are now certified to deliver Experience-Based Acceleration (EBA).
GoML is one of them.
GoML has a history of being among the launch partner for a number of strategic initiatives for AWS, particularly in Gen AI development. This is one more milestone and we look forward to helping our 120+ customers accelerate AI innovation.
Peter DeSantis: AI fuels unchanging cloud principles
The session centered on how the AI transformation is reshaping cloud infrastructure while reinforcing AWS’s long-standing principles security, reliability, performance, scalability, and cost-effectiveness.
Core AWS Principles (Unchanged Values):
- Security: Still the top priority as AI increases both opportunities and vulnerabilities.
- Availability and Performance: AWS continues to scale infrastructure to handle the most demanding workloads of the decade.
- Elasticity: A major strength; AWS aims to extend S3-like scalability to AI workloads eliminating capacity planning.
AI-Driven Infrastructure Innovation:
- Rapid growth in AI-driven applications is driving demand for highly efficient, adaptive cloud infrastructure.
- AWS has been preparing through decades of technology investments focusing on scalability and elasticity.
- Expansion of data centers and compute capacity is underway to support exponential AI demand.
Cost and Efficiency Focus:
- AI workloads are inherently expensive both training and inference are resource-intensive.
- AWS is working to reduce cost per inference and increase efficiency, making AI production-ready for more customers.
Strategic Direction:
- AWS aims to offer AI elasticity, letting developers scale models up or down seamlessly across regions and workloads.
- Expect tighter integration of AI within AWS services like S3, EC2, and Bedrock, enabling end-to-end, scalable AI solutions.
Deepak Singh: Agents revolutionize software building
Deepak Singh, VP of Developer Agents and Experiences at AWS, discussed how AI-driven agents are reshaping the software development lifecycle.
Development has evolved from simple code-completion tools to AI agents that can:
- Write complete, multi-file applications.
- Plan, reason, debug, and refactor projects.
- Work collaboratively with developers using natural language or diagrams.
- Complex tasks that once took months now take weeks or hours.
AI as a Collaborative Partner:
- These agents act as a “second brain” continuously learning from code, commits, and deployments.
- They recognize patterns, suggest improvements, and accelerate developer flow.
- The result: developers innovate faster and spend less time on repetitive, manual work.
Analogy and Reflection:
- Singh compared this AI moment to the launch of EC2 in 2006, when AWS first empowered developers through scalable cloud access calling today’s era equally transformative for AI empowerment.
AWS Mission for Developers:
- Empower every developer to create at the speed of thought using generative and agentic AI.
- Remove friction between vision and execution by automating tedious parts of building and deployment.
- Help builders focus on creativity, innovation, and meaningful work.
Strategic Outlook:
- AWS will continue integrating AI deeply into developer tools and workflows.
- The long-term vision: a seamless, AI-assisted development environment that boosts both productivity and fulfilment.
The Werner Vogels closing keynote
Dr. Werner Vogels, Amazon.com VP and CTO, delivered his 14th and final re:Invent keynote on December 5, 2025, stepping back after 13+ years to amplify younger AWS voices while staying at Amazon.
Developer Evolution and AI's Role
Vogels recapped 50+ years of shifts from COBOL/punch cards, assembly/compilers (1960s-80s), structured/OOP programming, Amazon's 1990s monolith-to-microservices pivot, to cloud's end of hardware planning stressing tools evolve but human work remains core. He addressed "Will AI take my job?" with "No if you evolve," citing Amazon teams thriving via adaptation, as AI accelerates progress across fields without replacing builders.
The Renaissance Developer vision
Vogels unveiled the Renaissance Developer like Da Vinci in a post-"dark ages" era of curiosity-driven invention equipped for AI's "Golden Age" (echoing Bezos).
- Curious: Drives learning, experimentation, and failure-tolerant innovation.
- Lifelong Learner: Socially via code reviews, travel/customer visits (e.g., Vogels' Africa/South America trips), and discomfort zones.
- Systems Thinker: Understands interconnected impacts (Yellowstone wolves analogy: structure/feedback changes outcomes).
- Communicator: Human-to-human and human-to-machine; specs reduce ambiguity (demo with Clare Liguori on spec-driven dev leading to Kiro tool).
- Owner: Owns software quality, avoiding "vibe coding" pitfalls.
- Polymath/T-Shaped: Broad knowledge + deep expertise, solving real-world issues like healthcare/nutrition.
Key challenges and progress
Vogels highlighted dev pain points like verification debt and AI hallucination, noting gains via automated reasoning/testing and spec-driven development. Code reviews transfer knowledge from seniors to juniors, growing future builders
If you missed them, our recaps have you covered for AWS re:Invent Day 1, AWS re:Invent Day 2 and AWS re:Invent Day 3.
For the Gen AI, here are some of the biggest announcements from AWS reInvent and our own guides:
Guide to Bedrock AgentCore
Synthetic ML data generation with AWS Clean Rooms
CloudWatch for agent observability
If you want a full rundown of all the big news from reInvent, you should head over to this AWS blog.
