AWS re:Invent 2025: The "Agentic" Era
re:Invent 2025 read through an agentic lens: the Nova 2 family split into specialised roles, and AWS stopped shipping chat and started shipping work.

If 2024 was about talking to LLMs, re:Invent 2025 was about letting them actually do the work. Here is the comprehensive breakdown of the most significant announcements.
1. The Amazon Nova 2 Model Family
AWS didn't just update their models; they built a specialized fleet for different agentic roles:
- Nova 2 Lite: Optimized for speed and cost. Equal or better than Gemini Flash 2.5 on 14/18 benchmarks.
- Nova 2 Pro: The "Reasoning" heavy-lifter. Best for complex multi-step tasks and long-range planning.
- Nova 2 Sonic: A speech-to-speech model for low-latency conversational AI.
- Nova 2 Omni: The true multimodal star. It processes text, images, video, and speech simultaneously with a 1M token context window.
- Nova Act: Generally Available and purpose-built for UI automation (browser-based tasks) with >90% reliability.
2. Custom Silicon: Graviton5 & Trainium3
The hardware story was about decoupling performance from cost:
- Graviton5: 25% faster than Graviton4, with 192 cores and a 5x larger L3 cache. It introduces the Nitro Isolation Engine, using formal verification to provide mathematical proof of workload isolation.
- Trainium3 & Trainium4: AWS announced Trn3 UltraServers and teased Trainium4, focusing on doubling the density of AI compute.
3. Frontier Agents: Your New Virtual Team
AWS launched three specialized "Frontier Agents" that work autonomously for days:
- Kiro: An autonomous developer that learns your codebase and handles features from end-to-end.
- AWS Security Agent: A 24/7 proactive security consultant that performs design reviews and automated penetration testing.
- AWS DevOps Agent: An "Autonomous SRE" that resolves incidents and monitors application resilience proactively.
4. Infrastructure & Database Wins
- Database Savings Plans: One flexible commitment for RDS, Aurora, DynamoDB, and more.
- P6e-GB300 UltraServers: Generally Available, offering 1.5x GPU memory and FP4 compute for the heaviest inference tasks.
- S3 Batch Operations: Now 10x faster, making massive data migrations nearly instantaneous.
More from Ercan
Two more sites, same author, different ground.
Cloud, AWS, EKS, Terraform, platform engineering.
Field notes from production systems. EKS, IAM, Terraform at organization scale, observability, cost optimization.
Visit ercan.cloud →The hub. About, consulting, contact.
Personal hub for both writing tracks. Who I am, how the consulting works, how to reach me.
Visit ercanermis.com →