Kaltura has entered into a definitive agreement to acquire eSelf.ai, an Israeli company that specializes in conversational avatars, for about $27 million. The acquisition brings eSelf’s high-fidelity digital humans, real-time speech capabilities, and screen-aware interaction technology to Kaltura’s enterprise video stack.
eSelf was launched in 2023 by CEO Alan Bekker, who, like Shoshan, also previously founded startup Voca (which was acquired by Snap in 2020), with CTO Eylon Shoshan. The eSelf team and founders will all be joining Kaltura to integrate the platform’s avatar studio and multilingual engine (supporting over 30 languages) into Kaltura’s product offerings.
Why Kaltura Is Betting on Conversational Avatars
For years, Kaltura made a name for itself as one of the first to focus heavily on video infrastructure and services—hosting, streaming, and interactive experiences—for enterprises and universities.
Video becomes a two-way interface next. With eSelf, Kaltura wants to match realistic avatars with corporate knowledge so AI agents can see, listen, and speak in real time and provide guidance instead of just playing content.
The strategic rationale is clear: context-aware, personalized AI agents have been proven to increase sales, onboarding, customer care, and training—spaces in which Kaltura already has scale. Rather than provide a “face” that rides on top of a chatbot, Kaltura hopes to ship the complete workflow—avatar, intelligence, and integrations—baked into your existing video portal, webinar solutions (like Zoom), virtual classroom tools, and media loaders.
What eSelf brings under the bonnet for Kaltura
eSelf’s stack ties together speech-to-video generation, low-latency speech recognition, and “screen understanding,” which allows an agent to understand what a user is looking at and respond accordingly. In other words, an avatar can view a product demo with you, interject as questions present themselves, and shape explanations to your role or level of expertise.
It also comes with a studio to design and customize photorealistic presenters, tune voices, and deploy across channels while preserving brand and compliance controls. That control actually matters in regulated environments; businesses are finding themselves needing auditability, usage policies, and content provenance signals as they start using synthetic media.
How it fits into Kaltura’s existing product offerings
Kaltura is the leading video technology provider in education that empowers any organization to create value from video. With its clients ranging from software companies to major technology, financial, healthcare, and education companies in the U.S. and globally, it has a customer list of more than 800 enterprises.
In-product, embedded eSelf agents might be making a sandwich inside a product video explaining how to use your food processor, serving as an always-handy 24/7 support rep that can “see” the customer’s screen with their permission, or playing the role of teaching assistant within an online university course.
Kaltura intends to deliver standalone agents that can be embedded in education, media and telecom, e-commerce, financial services, and even healthcare and pharmaceuticals—sectors where there is already significant grounding for video in the user journey.
Competitive Landscape And Differentiators
Avatar platforms in the digital realm are proliferating, with players including Synthesia, HeyGen, and Soul Machines, and tools developed on top of cloud AI speech and vision services. NVIDIA’s ACE initiative and big cloud service provider enterprise assistants are also moving in the direction of real-time, embodied AI experiences.
Kaltura’s advantage isn’t in novel avatars, after all, but in distribution and integration. The enterprise video workflow—from creation to analytics—already includes it, so avatars can plug directly into the Webex content libraries, identity systems, and data governance you may already be using. This positioning is key: Both Gartner and Forrester have stressed the point that integration, latency, and governance determine whether enterprises adopt AI from a commercial vendor (or not), and where they see return on investment, more than flashy demos.
Ultimately, success will be determined by measurability, like increased conversion on product pages, quicker support resolution, and higher learner completion rates. Look to Kaltura to talk about latency objectives, accuracy across multiple languages, and enterprise controls, including role-based access, content watermarking aligned with C2PA recommendations, and compliance with frameworks such as the NIST AI Risk Management Framework.
Deal context and what to watch after the acquisition
This will be its fourth acquisition, after Tvinci in 2014, Rapt Media in 2018, and Newrow last year. The company, which went public in 2021, is projected to report around $180 million in revenue for the year, is profitable on an adjusted EBITDA and cash flow basis, and has approximately 600 employees. eSelf last announced a $4.5 million funding round toward the end of 2024.
Short term, expect a roadmap that will integrate eSelf’s avatars with Kaltura’s webinar and learning products first, then move on to customer support and sales enablement. Regulated industries will be keeping an eye on what the pricing and deployment models—on-premises, private cloud, or multi-tenant—are for Azure services. Look for enhanced language support, better low-bandwidth lip-sync quality, and more seamless CRM and knowledge-base integration.
The larger view is that video is shifting from a passive form of content into an interactive interface. With eSelf, Kaltura is betting that the next generation of enterprise engagement is led by AI in a conversationally driven, screen-aware experience that looks and feels human—whilst ensuring compliance and control needs are met.
