Behind the launch of AIRA Agentic Marketing Suite: David’s reflection on the project and how is Kentico making AI work for marketers

Like in any role, there are many different situations to handle. The most challenging part is looking ahead and trying to anticipate where the market is heading and where we want to take the product so it makes sense for both our customers and partners. It’s about understanding how the market will evolve and then shaping our product strategy for Xperience by Kentico so it aligns with that direction. At the same time, it’s important not to blindly follow trends. We need to find a healthy balance between building attractive features and delivering things our customers genuinely need for their everyday work. 

Another big challenge is prioritization. Not everything we come up with can be delivered quickly. That means saying “no” quite often, even to ideas I personally find exciting. You have to rely on prioritization frameworks and make tough calls about what truly matters most.

What makes the role rewarding is when we get it right. When we manage to choose the right direction, design the solution well, deliver it to customers and they’re genuinely excited about it and tell us so. When you see that something the team struggled with, debated over, and refined together is actually helping people and they tell you.  

Kentico recently introduced AIRA Agentic Marketing Suite. Why is this such a significant milestone for the company?

We started working with AI quite early, but initially, it was more about experimentation. We focused on isolated use cases, improving specific workflows, optimizing images, or helping a bit with content creation. 

AIRA Agentic Marketing Suite is a well‑thought‑out, cohesive concept built for a new era where AI plays a fundamental role. It’s designed to take over repetitive, predictable tasks: especially the ones people don’t really want to do anymore.  

The key difference is usability. This isn’t about casually chatting with an AI and getting generic ideas. AIRA Agentic Marketing Suite was designed to work with the right context, so its outputs are relevant, accurate, and actually move users toward the result they expect. That required a lot of thinking and iteration and making sure the AI understands context and avoids vague, misleading, or low‑quality answers.

This is why AIRA Agentic Marketing Suite is such an important milestone for Kentico. For the first time, we’re presenting the market with a complete, usable solution that customers can start working with today and that we can continue to evolve and expand over time. 

How does AIRA Agentic Marketing Suite fit into Kentico’s long-term product vision?

Our long‑term product vision is to give marketing leaders a centralized view of what’s happening across all their channels. We often describe it as a single pane of glass; a place where leaders can see, manage, and understand everything in one place.

Within their teams, marketers, editors, and specialists all have different roles and responsibilities, and they need to collaborate effectively. In reality, teams aren’t always perfectly staffed. Sometimes roles are missing, people lack time, specific expertise, or space to experiment. This is exactly where AI agents add value. 

They can support teams by taking on repetitive or time‑consuming tasks. Agents don’t mind doing more work or working “overtime.” As a result, marketers can achieve significantly more in the same amount of time. AIRA Agentic Marketing Suite doesn’t replace teams, it complements them and makes them more effective. 

Introducing AIRA Marketing Suite.

AIRA Agentic Marketing Suite is built to introduce AI agents that can collaborate with marketing teams and in some cases fill gaps where specific roles or capacities are missing. These agents operate much like real team members. For example, the Content Strategist Agent automates content governance by auditing content and content models against brand standards.

The key value is a combination of expertise, consistency, and endurance. Agents can work continuously, follow established processes precisely, and always respect company workflows and approvals. What’s especially powerful is their ability to continuously work with context. They can reference marketing strategies, content plans, brand guidelines, or other key documents and consistently apply them across everything that’s created.

Previously, this required deep individual knowledge or complex review workflows involving multiple colleagues. Now, that context is always present. When someone asks an agent to adjust a paragraph, generate a description, or expand content for another channel, the agent automatically takes all relevant context into account.

In some cases, agents can even be more consistent than humans; not because people aren’t capable, but because humans naturally miss things over time. Agents don’t. They apply the same rules, context, and reasoning every single time. 

AIRA didn’t start as a single, big concept. It evolved over time. Initially, it was a simple AI assistant, answering questions, helping with documentation, or providing general guidance, similar to an early version of ChatGPT but without internet access. 

Over time, we added more AI-powered capabilities in the background: image variants, email subject suggestions, translations, and other features that quietly improved efficiency without users even thinking about AI. The challenge was figuring out how to connect all of this into a coherent concept, not just from a user experience perspective, but also from a product and development standpoint. We didn’t want isolated AI features anymore. We wanted AI to be part of how the product is designed and built going forward. 

That meant rethinking how we design agents. Each agent represents a role; with a purpose, responsibilities, and context. To do that well, we had to deeply understand how marketers actually work: what they think about, what decisions they make, what information they need, and where they struggle.

We conducted exploratory research with users, walked through real workflows, and asked detailed questions about how they evaluate success, analyze results, and make decisions. From there, we modeled agents that could support those processes end‑to‑end. My role was shaping the overall concept, making sure the agents aligned with real user needs, and ensuring we weren’t just building features but a system that can scale and evolve. 

What stood out to you about the team you worked with on AIRA Agentic Marketing Suite? 

The initial foundations were built by a Prague-based team focused on AI infrastructure. They laid the groundwork and showed how the system could function behind the scenes. But very quickly, it became clear that AI couldn’t live within a single team. Our strategy is that everyone at Kentico needs to understand and work with AI; especially teams closer to customers and real marketing use cases.

Today, those teams bring deep domain expertise and actively shape how agents should behave in real-world scenarios. The original AI team now acts more as an enabler, supporting others when something new is needed. 

That shift was crucial. It led to shared knowledge, common documents used by multiple agents, and a much stronger, company-wide approach to AI.

Defining Moments That Shaped AIRA.

Every major project comes with challenges and highlights. Can you share a few defining moments that really shaped the outcome?

One of the biggest challenges was coordination. The more teams involved, the more complex planning becomes. Dependencies, timing, and alignment suddenly matter much more. 

Another challenge was mindset. We were used to building predictable, long-term platform features. AI changes that completely. The market evolves weekly, and we had to constantly reassess whether our thinking still matched reality.

One major highlight came during early prototype testing with customers. Even before the product was finished, users reacted with genuine excitement. They could ask questions directly about their own websites or data and immediately get meaningful explanations and recommendations. Seeing customers naturally explore, ask follow-up questions, and connect the dots themselves was a huge validation. Even at the prototype stage, it was clear that the potential was real.