AWS Summit NYC - Recap

Introduction

At this year’s AWS Summit NYC, the energy inside the Javits Center felt like a collective realization: the future of AI is not only imminent, but rather foreseeable, autonomous, and happening faster than anyone expected. With over 200 sessions and deep dives across industries, it was clear AWS is charting the next phase of cloud intelligence. And leading that charge? AWS AgentCore, a brand-new service that might just redefine how businesses operationalize AI agents.

Agentic AI Takes Center Stage

After attending the AWS Summit in DC in June, we published a blog post titled  AWS Summit 2025 Reflection: AI Agents Are Ready. Are You? Turns out, they were more ready than any of us expected. Swami Sivasubramanian, AWS VP for Agentic AI, led the keynote in NYC on Wednesday, unveiling the next evolution in autonomous systems. Introduced during Sivasubramanian’s keynote, AWS AgentCore represents a foundational leap for agentic AI. He discussed how agents should not only chat, but act: booking trips, optimizing supply chains, even managing reservations.  With AgentCore, developers can construct AI agents that:

• Interface with enterprise systems (ERPs, CRMs, ticketing systems)

• Handle complex workflows like travel booking, inventory forecasting, or IT ticket triage

• Integrate seamlessly into Amazon Bedrock with tools like Amazon Q Apps, making agent behavior modifiable and governed via reusable knowledge bases and role-based controls

And most notably, AgentCore is launching with support for Anthropic Claude, Meta Llama 3, and Amazon Titan, offering developers maximum flexibility on model strategy.

Why AgentCore Matters Now

We’re at a turning point in AI adoption. Generative models gave us prototypes and demos. Agentic AI, with the infrastructure provided by AgentCore, gives us products. It's a go-to-market platform for agents that solve business problems autonomously and at scale. 

There’s no avoiding the question dominating headlines: “Is AI going to take people’s jobs?” It’s a valid concern and one that deserves clarity. What we’re seeing with AgentCore and agentic AI isn’t about replacing humans wholesale. It’s about augmenting human capability and relieving knowledge workers of the digital busywork that’s consumed too much of their time. These agents are not creative directors, decision-makers, or strategic thinkers. They excel at what slows most teams down: navigating APIs, bridging data silos, orchestrating multi-step workflows, and transforming messy input into structured, actionable insight. These are the connective tasks that eat up time, and agentic AI can handle them, at scale and on demand.

Agentic AI, when designed responsibly, actually elevates human roles by:

• Automating the repetitive scaffolding work around decision-making

• Freeing up time for strategic thinking and client engagement

• Creating a foundation for faster experimentation and iteration

Just as cloud computing redefined IT roles (without eliminating them), agentic AI is likely to redefine knowledge work by expanding it, not erasing it. And platforms like AgentCore give businesses the structure and confidence to pursue this transformation deliberately, with oversight, guardrails, and impact metrics in hand.

Agentic AI at Red Oak

At Red Oak Strategic, we are weaving Agentic AI into the fabric of how our clients operate their businesses. Our team is building AI agents that connect to real business systems, reason through complex workflows, and take meaningful action across cloud platforms, data pipelines, and analytics environments. As an AWS Advanced Tier Partner, we of course use Amazon Web Services to ensure our solutions are both scalable and governed, delivering true operational value. Our goal is to help clients move beyond the hype and put autonomous systems to work where it matters most. For example, we’re deploying a custom agentic AI solution for one client to automate proposal and document review: the agent ingests vendor proposals, analyzes them against business criteria, and generates a structured summary of potential risks, turning what was once hours of manual review into minutes of intelligent insight.

Final Thoughts

AWS Summit NYC confirmed what many of us in the space already sensed: we’re past the prototype era. The tools are mature, the frameworks are hardened, and the workflows are agent-ready. AWS AgentCore may be new, but its implications are profound, and we’re not just watching from the sidelines. We’re building, testing, and preparing to scale. If you’re thinking about AI agents in your enterprise, let’s talk. The next generation of automation is already under way.