Almost exactly a year ago, we interviewed Mike Pirello, CEO of ePublishing, the provider of an integrated tech stack built specifically for the media industry. In today’s media world, with a frenetic pace of change, a year is a long time indeed. Media companies are focused on strengthening their bond with audiences, expanding those audiences, and building a core collection of the technologies that enable those things.
In short, it’s the tech stack, and it’s at the center of all that’s happening in media. We believe ePublishing is too. So when we had the opportunity to follow up on our prior interview with Pirello, we jumped at the chance. We sat down with Kelli Chmielorz, ePublishing’s vice president of products, and she caught us up on the changes that have occurred over the last 12 months. Here’s our interview.
Fox Tales: As head of product for media companies, what is the specialized media market (B2B and enthusiast B2C) prioritizing these days?
Kelli Chmielorz: The priorities have converged around a few things that felt aspirational two years ago and are now table stakes. First is revenue diversification. Specialized publishers have realized that ad dependency is a structural vulnerability, not just a cyclical one. So you’re seeing a real push into subscriptions, membership tiers, events, e-commerce, and data licensing. Second is first-party data. With third-party cookies effectively gone, publishers who built direct audience relationships are winning, and everyone else is scrambling to catch up. Third—and this is where I spend most of my time—is operational efficiency. B2B and enthusiast publishers tend to run lean. They can’t afford the bloated tech stacks that enterprise-media companies carry, so there’s enormous pressure to do more with less, while still competing for audience attention against platforms that have infinite engineering resources.
Fox Tales: How much is AI changing your product mix?
Kelli Chmielorz.
Chmielorz: Substantially. And faster than I expected even 18 months ago. AI has touched almost every layer of what we offer. On the content side, we’ve built AI into editorial and design workflows to help teams produce more without sacrificing the subject-matter depth that defines specialized media. On the audience side, our Ask My Brand product is a conversational AI search layer that sits on top of a publisher’s content—and critically, it’s built to serve ads and drive subscriptions, not just answer questions. Ask My Brand is a very unique product that is being adopted incredibly quickly, that simply didn’t exist in our portfolio two years ago.
The newest frontier for us is autonomous agentic infrastructure. Our MediaFabric Gateway is an MCP server—the first one purpose-built for publishers—that turns a publisher’s CMS into a live context layer accessible to AI agents. Something we call “liquid context.” That’s a fundamentally different kind of product than anything we’ve offered before. We’re moving from helping publishers manage content to helping publishers make their content machine-readable and monetizable in the AI economy.
Fox Tales: Your tagline is “The Modern Media Tech Stack.” How would you define that, and why?
Chmielorz: The “modern” in that phrase is doing real work. For too long, the media tech stack meant a CMS bolted to a separate subscription platform, bolted to a separate analytics tool, bolted to a separate ad server—each with its own data model and none of them talking to each other cleanly. That’s not a stack, that’s a junk drawer.
When we say Modern Media Tech Stack, we mean a purpose-built, integrated system where content management, audience operations, commerce, and AI live in a single architecture sharing a single data layer. The practical result is that editorial, circulation, and revenue teams are all working from the same picture of the business. No reconciliation, no data lag, no “which number is right?” conversations. Modern also means AI-native—not AI bolted on, but AI embedded in the core workflows and the infrastructure itself. For a specialized publisher competing for audience attention in 2026, that’s not optional anymore.
Fox Tales: What’s around the corner in media that operators might not be seeing yet?
Chmielorz: The biggest thing most operators are underestimating is the shift from search-driven traffic to agent-driven content discovery. Right now, a lot of publishers are still optimizing for Google. But the traffic model is fundamentally changing—AI agents and assistants are increasingly answering questions directly rather than sending users to publisher sites. The publishers who will win are the ones whose content is accessible, citable, and monetizable within that AI layer, not the ones hoping the old referral model holds.
The second thing is content licensing at scale. Specialized publishers sit on extraordinarily valuable, deeply verified IP—the kind of authoritative data that AI models need and generic web scraping can’t reliably provide. There’s a real business emerging around licensing structured content to AI systems, and most operators haven’t built the infrastructure to participate in it yet. That’s a significant missed opportunity.
Fox Tales: How important is the single, unified architecture that you referred to in the tech stack?
Chmielorz: It’s the difference between running a media business and managing a collection of software subscriptions. I’ve watched publishers spend enormous energy on integration work—trying to get their CMS to talk to their audience database to talk to their analytics to talk to their billing system—and that energy doesn’t produce a single dollar of reader revenue. It’s pure overhead.
A unified architecture matters for three reasons. The first is data fidelity—when every system shares one data model, you have one version of the truth about your audience, your content performance, and your revenue. The second is speed—product decisions that used to require a data warehouse query and a three-team coordination call become operational. The third, and most important right now, is AI readiness. AI works best when it has a complete, coherent context. A fragmented stack produces fragmented signals. A unified platform lets you deploy AI across content, audience, and commerce in a way that’s actually intelligent—because it’s seeing the whole business, not pieces of it. That’s the architecture we’ve built, and increasingly it’s what publishers are demanding rather than accepting.