By Paul Allen in strategic leverage — 11 Nov 2025

Kaltura Acquires eSelf for $27M to Embed Generative AI Avatars in Enterprise Video Workflows

Kaltura, a leading provider of video platform solutions for enterprises, acquired the AI avatar startup eSelf in a $27 million deal in November 2025. Founded by the creator of Snap’s AI capabilities, eSelf develops generative AI avatar technology that enables synthetic human representations tailored for enterprise video and learning tools. Kaltura’s business model centers on monetizing video hosting, streaming, and interactive learning services for businesses, and the acquisition will integrate these AI avatars deeply into Kaltura’s existing platform tools to enhance engagement and personalization.

Embedding AI Avatars Targets the Enterprise Video Engagement Bottleneck

Kaltura operates in a market where video content is abundant but user engagement, especially in corporate training and communications, plateaus due to static, one-way video formats. By acquiring eSelf, Kaltura is addressing this fundamental constraint: the lack of interactive, personalized video experiences that can scale without dramatically increasing content production costs or relying on live presenters.

eSelf’s generative AI avatars automatically create realistic synthetic humans that can be scripted or respond dynamically, positioning Kaltura to offer enterprises video tools that automate personalization at scale. Unlike alternatives that rely on prerecorded hosts or require hiring actors, Kaltura can generate on-demand avatars across languages and personas without ongoing human effort. This shifts the constraint from expensive, manual content creation to AI-driven scalability.

Leveraging Founder's Snap AI Expertise to Accelerate Avatar Realism and Interactivity

eSelf’s founder, previously responsible for Snap’s AI capabilities, brings experience building consumer-grade, scalable AI systems optimized for real-time performance on mass platforms. This differentiates Kaltura’s acquisition from generic AI avatar startups by infusing consumer-level interaction quality into enterprise tools. For example, Snap’s AI technology enabled over 500 million users to manipulate realistic avatars and lenses daily, reflecting operational maturity in avatar rendering and responsiveness.

Integrating eSelf’s models means Kaltura can potentially embed avatars that lip-sync, gesture, and adapt dialogue in real time within corporate videos or learning modules—a leap beyond static video or simple chatbots. This integration promises to reduce video content production cycles from weeks to hours by automating presenter creation. It also opens the door to leveraging AI avatars as always-available virtual trainers or customer assistants embedded directly into enterprise video workflows.

Choosing In-House AI Integration Over Outsourcing to SaaS AI Providers

Instead of integrating generic third-party generative AI services, Kaltura’s acquisition internalizes avatar AI capabilities, capturing control over the technology stack and data flow. This design improves system-level leverage by:

Reducing per-use AI costs by owning training data and model updates, rather than paying per video or viewer through external APIs.
Enabling tight embedding of avatars within Kaltura’s proprietary video platform, which preserves user experience continuity and simplifies enterprise deployment.
Giving Kaltura the ability to customize avatars extensively for vertical markets (e.g., compliance training, sales demos), which off-the-shelf AI providers currently lack.

By controlling the AI tooling, Kaltura avoids fragmented vendor lock-in and can innovate faster on avatar features, such as synchronizing avatars with interactive video quizzes or real-time feedback systems.

Integrating Generative AI Alters Kaltura’s Core Product Constraint From Content Creation To User Interaction

Traditionally, Kaltura’s growth has depended on expanding video hosting capabilities and scaling content delivery networks. The addition of eSelf’s generative AI avatars repositions Kaltura’s bottleneck from capacity to engagement. Customer value will derive less from raw content volume and more from AI-driven interactive elements that increase learning retention, sales conversions, or internal communications reach.

This leverages Kaltura’s existing installed base — tens of thousands of enterprise customers worldwide — by embedding AI at the point of video consumption rather than acquisition or production. A concrete example: an enterprise training video powered by AI avatars can dynamically adjust script complexity in real time based on viewer responses, a feature nearly impossible without integrated generative AI.

Comparison With Industry Alternatives Highlights Unique Strategic Position

Companies like Synthesia and Hour One offer AI avatar SaaS solutions but lack deep integration with enterprise video systems. For instance, Synthesia charges approximately $30 per video minute, which can scale to high costs for large user bases. Kaltura’s internal AI capabilities could drop incremental avatar deployment costs to infrastructure levels, turning this from a variable cost operation into a fixed-capacity system.

Unlike generic chatbots, eSelf’s avatar technology prioritizes realism and rich media context, which matter in high-stakes enterprise training and communications where trust and engagement are crucial. This contrasts with large language model-based avatars that work mainly as text interfaces.

Extending Kaltura’s Systemic Advantage Through AI-Driven Content Personalization

The acquisition provides a leverage point for Kaltura to build an internal feedback system: AI avatars can collect granular interaction metrics on viewer engagement within videos, feeding into machine learning models that optimize content sequencing automatically. This creates a self-improving system with minimal human input. Over time, this shifts costly human content managers to strategic overseers rather than production workers.

This strategy aligns with broader industry shifts where AI integration transforms core product constraints — similar to how Google’s Gemini AI augmented TV voice interaction or how AI tools automated scaling side hustles by eliminating user acquisition and operations bottlenecks. Kaltura’s move pinpoints engagement as the next frontier instead of plain video delivery or hosting.

Kaltura’s integration of AI avatars to transform enterprise learning underscores the growing importance of engaging and scalable e-learning platforms. If you're looking to enhance your own training programs with dynamic, interactive content, platforms like Learnworlds provide a powerful foundation for creating and selling online courses tailored to today’s digital learners. This is exactly why Learnworlds has become essential for educators and businesses aiming to elevate learning experiences through technology. Learn more about Learnworlds →

💡 Full Transparency: Some links in this article are affiliate partnerships. If you find value in the tools we recommend and decide to try them, we may earn a commission at no extra cost to you. We only recommend tools that align with the strategic thinking we share here. Think of it as supporting independent business analysis while discovering leverage in your own operations.

Frequently Asked Questions

What are generative AI avatars and how do they improve enterprise video workflows?

Generative AI avatars are synthetic human representations created using AI that enable dynamic and personalized video content. They improve enterprise video workflows by automating presenter creation, enabling real-time interaction, and reducing production times from weeks to hours while enhancing engagement and personalization.

How can integrating AI avatars reduce costs compared to traditional video production?

AI avatars eliminate the need for hiring actors or prerecorded hosts, allowing enterprises to generate on-demand avatars across languages and personas without ongoing human effort. This reduces content creation costs significantly and transforms variable per-use costs into fixed infrastructure costs.

What makes Kaltura's AI avatar integration different from other SaaS AI providers?

Kaltura internalizes AI avatar capabilities by owning training data and tightly embedding avatars into its proprietary platform. This approach reduces per-use costs, avoids vendor lock-in, and enables extensive customization for vertical markets, whereas SaaS providers typically charge up to $30 per video minute without deep system integration.

Who founded eSelf and why is their expertise significant?

eSelf was founded by the creator of Snap’s AI capabilities, bringing experience in consumer-grade, scalable AI systems that support real-time avatar rendering for over 500 million users daily. This expertise allows Kaltura to offer high-quality, interactive avatars with realism and responsiveness beyond typical enterprise tools.

How does integrating AI avatars shift Kaltura's core product constraint?

By embedding generative AI avatars, Kaltura shifts its bottleneck from video content creation capacity to user engagement. The AI-driven interactive elements increase learning retention and communication reach, moving value focus from content volume to personalized video experiences.

What are some practical applications of AI avatars in enterprise videos?

AI avatars can dynamically adjust script complexity based on viewer responses during training, act as virtual trainers or customer assistants, and synchronize with interactive video quizzes or feedback systems to enhance learning and communication outcomes.

How does Kaltura's AI avatar technology compare cost-wise to competitors like Synthesia?

Synthesia charges about $30 per video minute, which can lead to high costs at scale. Kaltura’s in-house AI capabilities can reduce incremental costs to infrastructure levels, making avatar deployment more cost-effective for large enterprise user bases.

What future benefits does Kaltura expect from AI avatars regarding content personalization?

AI avatars enable collecting detailed engagement metrics within videos, which support machine learning models that automatically optimize content sequencing. This creates a self-improving system that reduces human content management and drives strategic content enhancement over time.