Kaltura Agentic Digital Experience Manifesto

From available knowledge to fluent, 1:1 experiences.

Nobody taught you how to talk to AI. 

You ask, and when the answer misses, you do what you’ve always done – add context, cut in, ask for the short version. Nobody wrote you a manual for that. Conversation is the oldest skill you have. 

Every other piece of software made you learn its language first. A magnifying glass meant search. Three lines meant menu. Buying something meant a funnel with a checkout at the end. Learning something meant modules, in somebody else’s order. Good design made this easier to learn. 

The internet became extraordinarily good at making knowledge available. But availability was never the same as understanding, and understanding only becomes valuable when you can use it. At Kaltura, we call that fluency – the point at which knowledge has been learned deeply enough to inform a decision, improve performance, or enable action. 

That is what we’re building with our Conversational Agent Studio: agents that understand what an organization knows, build context about the person they’re interacting with, and use both to shape what happens next. As intent becomes clearer, they can surface the right knowledge, generate the entire interface around it, change the path through the journey, and take action toward a defined goal.  

The result is more than a conversation tailored to one person. It is an experience that can keep adapting until that person understands enough to decide, perform, learn, or act. 

From click browsing to conversational browsing 

The web was built around navigation: search, click, filter, browse, choose. The system could not know exactly what you wanted, so interfaces were designed to help you find your way toward it. 

AI changes that relationship. People can increasingly say what they need directly, clarify it as they go, and expect to get the answer, explanation, recommendation, demonstration, or action itself in the form that is most useful to them. Intent no longer has to be inferred entirely from clicks. 

But despite the transition from click browsing to conversational browsing, most digital experiences are still built before the person arrives. Thousands of users encounter the same homepage, course, product tour, or content path, with personalization layered onto a journey designed in advance. 

Conversational AI gives us a new way to move beyond that limitation. The combination of expressed intent and observed behavior in a conversation interface provides the system with a much richer understanding of the person in front of it: what they have viewed, returned to, ignored, or acted on, alongside what they are now explicitly trying to understand or accomplish. 

However, the bigger shift is that this understanding can shape the experience itself in real time. What a buyer sees, how a learner is taught, what action a customer is offered, and even the path through the experience can change as the system understands more. 

Until now, the interface was largely designed first, and the person moved through it. With dynamic, generative UI, the interface becomes part of the interaction itself. A user asks for a comparison, and the comparison appears. They need a demonstration and the relevant video or visual is generated in real time. They reveal a gap in understanding and the explanation, sequence, or format changes. They are ready to act, and the appropriate workflow or next step appears in context. 

The interface is no longer just the container for the conversation. It becomes something the agent can generate and adapt around the person in real time. 

Much of the current agent category concentrates on the conversation itself: how naturally a voice listens, how convincingly a face responds, how quickly the system can render something useful back. Those are hard technological problems, and solving them matters. But the larger opportunity begins once the conversation can reshape everything around it. 

Fluent, not just informed 

Kaltura has spent two decades working on the distance between available knowledge and useful understanding. 

Video libraries, courses, event platforms, and learning environments may look like different categories, but they share an underlying job: take what an organization knows and help another person understand enough of it to do something. We built systems for recording that knowledge, structuring it, teaching with it, and understanding how people engage with it. Conversational AI brings those capabilities together in a new experience model. 

Conversational AI doesn’t give us a new mission. It changes how directly we can pursue the one we already had. 

Knowledge is now actionable in a much more dynamic way. The agentic platform can decide what to explain, what to show, which expertise to apply, how to adapt the journey, and when to move from information into action. At the same time, every interaction adds to the system’s understanding of the person, creating more context for what should happen next. 

That is what we mean by fluency: not simply giving someone access to knowledge, but using knowledge, context, and interaction together to help them move forward. 

It is also the foundation of the next generation of Agentic Digital Experiences: experiences built on organizational knowledge, shaped by real-time understanding, and able to adapt, guide, and act around the person using them. 
 

Knowledge becomes capability 

For that experience to be useful, the agent needs more than access to a knowledge base. 

Organizations hold enormous bodies of knowledge across video, documents, systems, methodologies, and accumulated expertise. The challenge is understanding what is inside that knowledge, breaking it down into meaningful components, and knowing what can be surfaced, combined, taught with, demonstrated, or acted on in a particular moment. 

A recorded demo contains product expertise. A course contains a teaching model. A webinar contains arguments, examples, questions, and explanations. A methodology contains a way of working. 

When the system understands that material at the level of meaning and structure, the expertise inside it becomes something an agent can use. 

This is Content → Skill → Agent: content carries the knowledge, content intelligence turns the expertise inside it into usable skills, and the agent applies those skills toward a goal. 

Agentic Digital Experiences: Our approach 

Mission-driven agents 
Every agent starts with a defined goal: teach, guide, qualify, advise, resolve, or convert. 

Content intelligence 
Agents understand the meaning and structure of organizational knowledge so they can surface, combine, explain, demonstrate, or act on it. 

Compounding context 
Identity, history, behavior, engagement, and conversation build a richer understanding of the person and their intent. 

Content → Skill → Agent 
Content intelligence turns organizational expertise into skills an agent can apply toward a goal. 

Generated journeys and Generative UI 
The content, interface, sequence, and next step can all change as the interaction develops. 

Fluency 
Success is measured by how far the experience moves someone from information to understanding, capability, and action. 

What you can build today 

The Conversational Agent Studio puts this experience model directly in organizations’ hands. 

Build an agent around a business goal. Ground it in your video, content, and organizational knowledge. Define its persona, behavior, rules, actions, and brand. Give it an avatar, voice, or text interface. Then deploy it into customer, employee, and learning journeys. 

The agent brings together what your organization knows, what it already understands about the person, what it learns through the conversation, and the skills and actions available to it. From there, it can generate what should happen next – the answer, the content, the interface, the path, or the action. 

A product library can advise. A methodology can coach. A training program can teach around a gap. Knowledge becomes reusable agent capability. 

Every interaction can leave both sides knowing more: the user more fluent in what they came to understand or accomplish, and the organization better able to anticipate what they need next. 

The internet made knowledge available. Kaltura is making it fluent. 

— Kaltura