Kaltura Ships Dynamic UI, Letting AI Agents Generate Interfaces Mid-Conversation
The Development
Kaltura, Inc. (Nasdaq: KLTR) announced general availability of Dynamic UI for its Conversational Agents. The capability lets AI avatars select, generate and assemble videos, presentations, animations, interactive galleries and quizzes inside the flow of a conversation, adapting as the user's questions and progress change. Agents draw on an organization's own content libraries, metadata and enterprise knowledge, within enterprise-defined controls. Kaltura cites use cases in learning, product exploration for prospective customers, TV entertainment guides, and interactive kiosks and digital signage. CEO Ron Yekutiel framed the release as a move toward experiences generated and continuously adapted around each individual, building on nearly two decades of enterprise video technology.
Our Take
Kaltura's pitch is enterprise-wide, but the marketing signal is narrower and more important: the unit of design is shifting from the page to the content library. When an agent assembles the experience, your assets, metadata and brand rules become the product. Weak tagging and thin video inventory will surface instantly as a worse experience. Kaltura's grounding claim also targets the main objection to generative interfaces, off-brand improvisation. Treat it as a vendor assertion until tested, but the direction is right. The durable advantage goes to teams with deep, well-structured, rights-cleared content, not teams with the best templates.
What Changed
Conversational agents can now compose the visual experience, not just the answer. Layout, media selection and interactive elements are assembled per user at runtime from a governed content base, replacing the pre-built page or journey as the container for brand experience.
Marketing Impact
Product marketing and demand gen gain a path to demos and comparisons assembled per prospect, replacing static decks and fixed landing pages. Creative teams shift from producing finished pages to producing modular, metadata-rich components. Marketing operations inherit new governance work: approved asset pools, guardrails, and logging of what each agent actually showed.
Competitive Implication
Organizations with large, well-tagged video and content libraries gain structural advantage, because the agent's output quality is capped by its inputs. Page-centric web teams and template-driven personalization vendors are exposed: if the interface is generated per conversation, A/B-tested layouts and segment-based journeys lose their reason to exist.
Strategic Outlook
Expect rivals across CMS, DXP and conversational AI to ship generative-interface features through 2027, making runtime-assembled experiences a baseline expectation in high-consideration sales and learning. Measurement is the open flaw: session and page-based analytics cannot attribute outcomes to experiences that differ for every visitor.
The Exploit
Action Item
Heads of content and marketing operations should audit their video and product-content libraries for metadata depth, rights clearance and brand-rule coverage before Q1 2027 planning closes, then pilot one agent-assembled product walkthrough against a static landing page.