Building the Agent Ecosystem: 

By Everett Zufelt, Program Lead, Agent Ecosystem, MACH Alliance

The Agent Ecosystem initiative isn't just a new program — it's a new kind of process. Here's what we've been building, and why we need you in the room.

About six weeks ago, I stepped into a secondment as Program Lead for the MACH Alliance's Agent Ecosystem initiative. In that time, I've had conversations with vendors, enterprise architects, brand leaders, digital practitioners, and Alliance members across the ecosystem. And the thing I keep hearing — regardless of role, company size, or how far along anyone is on their AI journey — is a version of the same honest admission:

"We don't have this figured out yet."

The Agent Ecosystem initiative exists precisely because we believe the enterprises, practitioners, and technology partners navigating the agentic landscape shouldn't have to figure it out in isolation. The future of enterprise AI will not be defined by a single agent, a single vendor, or a single platform. It will be built — collaboratively, iteratively, and with plenty of productive friction — by communities like this one.

From Vision to Charter

Over the past month, the work has been about translating a compelling vision into something executable. That meant sitting across from a wide range of stakeholders — from technical architects worried about reliability and observability, to brand leaders asking "where do I even start?" — and listening carefully to what they actually need.

Agent Ecosystem Vision

“To become the independent agent ecosystem for commerce, digital experience, and customer data.”

What emerged is a vision and charter organized around four strategic programs: engaging the right people, building shared awareness of what's real now versus what's still speculative, demonstrating working solutions, and creating credible trust signals the industry can point to.

It sounds tidy on paper. In practice, it's a live conversation that evolves every week.

That's intentional. We're not building a finished framework and then broadcasting it to the world. We're building it with the people who need to use it — which means the initiative is only as strong as the community participating in it.

What We're Building (And How to Get Involved Now)

The initiative is organized around four strategic themes: engaging the right people, building shared awareness of what's real today, demonstrating working solutions, and creating credible trust signals the industry can point to. Each theme has programs behind it — and several are already open for participation.

Show & Tell and Tech Showcase sessions are the lowest-barrier entry point. These are regular showcases where vendors share what they're actually building — not a sales pitch, but real discussions with real lessons. If you have something to show, or simply want to learn from those who do, this is the place to start.

The AI Exchange is where we collaboratively build, test, and pressure-test practical AI solutions grounded in the principles of open, composable, and connected systems. If you have a use case worth exploring — or learnings from production that others would benefit from — contributing to the AI Exchange is a direct way to shape what good looks like for the broader community.

Two working groups are now open. The Agent Adoption & Operations Working Group is focused on readiness: helping enterprises understand where to start, how to build operating models that hold up at scale, and what governance actually needs to look like in practice. The Enterprise Agent Architecture Working Group is for technical leaders focused on the patterns, shared vocabulary, and reference designs needed to run many agents reliably across many systems. 

More opportunities to participate will be announced in the coming weeks as we build toward MACH X: Toronto on April 28–29 — where much of this work will come together in person for the first time.

Two Audiences, One Shared Challenge

Throughout the charter process, one thing became very clear: the enterprise audience for this work isn't monolithic.

Brand executives and digital leaders need a clear, honest view of what's possible now — not the breathless pitch deck version, but the "here's what it actually takes to operate this reliably" version. They need examples they can bring into internal conversations, starting points that account for governance, and a realistic picture of the sequencing required to move from pilot to production without accumulating debt they'll spend years unwinding.

Technical leaders and enterprise architects need something different: reference models, interoperability patterns, and day-two operational guidance. They're not worried about whether agents are real — they're worried about bounded autonomy, fallback behavior, observability, and how to build systems that don't require heroic intervention every 90 days to stay functional.

The initiative is designed to serve both — not with one-size-fits-all outputs, but with content and artifacts calibrated to the specific job each person is trying to do.

An Honest Invitation

Here's what I've learned in nearly two months of conversations: the enterprises and practitioners who are moving well on agentic AI aren't the ones with the most sophisticated technology. They're the ones who are being honest about what they know, learning from each other, and resisting the temptation to treat every deployment as a proof-of-concept PR moment.

That spirit is what the Agent Ecosystem initiative is built for.

If you're a brand leader trying to make sense of where to start, the guidance we publish and solutions we showcase will produce decision-grade guidance built from real implementation experience — not theory. If you're a technology partner with hard-won learnings from production deployments, the AI Exchange and working groups are places to contribute those learnings and help shape what good looks like. If you're a practitioner somewhere in the middle — curious, cautious, and looking for a community that takes the "how" as seriously as the "what" — this is the room.

We don't have the full picture yet. Nobody does. But we're building the conditions to find it together — and that work is underway now.

Visit AgentEcosystem.org to learn more, and reach out if you're ready to get involved.