Scaling expertise, one avatar at a time 

Whether you are trying to close pipeline, train new hires, or onboard customers, the bottleneck is often the same: your best experts cannot be everywhere at once. A specialist who knows the product inside out simply does not have the time to run every session or jump on every call. 

Because an expert cannot be in dozens of places at once, organizations usually resort to static handoffs: detailed documentation, shared PDFs, or dense slide decks. Yet working through a deck takes considerable effort, and vital nuances get lost along the way. While video is proven to hold attention and communicate nuance far more effectively, pulling specialists away from their core responsibilities to script, record, and edit every time a message updates creates a production bottleneck of its own. Besides, not all experts are the best on-air presenters. 

Avatar Video solves that trade-off. It turns what your team already has: presentations, documents, recordings, into a finished, narrated video in minutes. Crucially, it also puts a presenter on screen. Decades of research in learning science show that people understand and retain information much more deeply when guided by a lifelike presenter, a phenomenon known as the persona effect. Viewers naturally respond to an avatar’s facial expressions and conversational delivery in much the same way they respond to a person on camera. It provides the cognitive and retention benefits of a face-to-face walkthrough, while allowing the expert behind the knowledge to remain focused on their daily work. 

What is Avatar Video and where can it lead? 

Avatar Video is Kaltura’s answer to that instinct: feed it a piece of source material and it produces a finished, narrated video, presenter, voice, visuals, and structure, built and delivered by a realistic avatar. It’s a complete deliverable on its own. For teams that want to go further, the same avatar can extend into a Video-to-Live Agent, so the moment a viewer finishes watching, that same presenter can pick up a live, two-way conversation grounded in the material the video came from. 

Three things make that combination worth paying attention to. It’s engaging, because a face holds attention. It enables scale, because one video can be produced, translated, and reused across audiences without a production team standing behind every version. And it enables personalization, because the same underlying knowledge can be delivered differently to a new hire, a customer, or a partner, without rebuilding the content from scratch each time. 

Why do subject matter experts need video-making capabilities? 

Here’s the problem we set out to solve for our enterprise customers: the person who best understands your product, who knows exactly why a customer churns, who’s run the compliance training a hundred times, in short, the expert is rarely the same person who knows how to shoot, edit, and publish a video. That means their expertise stays trapped in a document, a slide deck, or their own head, and it’s somebody else’s job to translate it into content secondhand, or more frequently it just never gets made. 

Avatar Video is built specifically for that person. A product manager, a trainer, an HR lead, a customer success rep – anyone can produce something that looks and sounds professional in the time it used to take to just schedule a shoot. 

That’s the shift worth getting excited about: expertise becoming video, only without the need for a video creator in the loop. The impact is clear: research on AI-assisted video editing shows teams can accelerate content production by up to 95 percent, while McKinsey’s analysis of generative AI’s economic potential suggests the same teams can produce three to five times more output by employing AI solutions. 

Can someone with zero production experience finish a video on their first try? 

Instead of dropping the user in front of a blank canvas Avatar Video starts from what you already have. There are four ways to begin: start from a script, generate a video from a presentation, turn a recorded session into a highlights recap, or point it at a topic and let it draft an explainer from your existing videos, documents, or a webpage.

All the user has to do is pick a path, preview the scenes it builds, adjust the script or the avatar if needed, and click generate. The pacing, framing, and visual pairing are handled for you, with AI-selected B-roll pulled in automatically to keep the narrative visual rather than a talking head for ten minutes straight.

Choosing who delivers the video is part of that same simple flow. Avatar Video provides a library of professional presenter avatars to pick from, a range of looks, tones, and delivery styles, so a piece of content can be matched to its audience without any casting decision beyond a menu. The same tool also lets someone build an avatar from their own photo and voice: upload a single photo and audio file, and Avatar Video generates a digital twin, a presenter who looks and sounds like you, ready to narrate any script the same way you would. That option matters most for the person a video is really about, a CEO delivering a quarterly update, a subject matter expert whose face is already the trusted one in the room.  

This matters most when teams need video but lack time or production support. A regional sales lead may need a quick, localized product update for a specific account but they don’t have the production budget or a week to spare waiting for an editor. A trainer who wants to turn their session into reusable content can’t ask their learners to wait for their turn in a media services queue. Ease of use is what turns a tool like this from a good idea into something people use.  

By removing the workflow dependencies of traditional video production, organizations have reported cutting production costs by 50 to 70 percent, according to strategy firm Bain & Company’s research on AI in content production. 

How one video become a hundred versions? 

Because the presenter, script structure, and visuals aren’t baked into a single file the way they are in a traditional edit, the same core video can become a different cut for a different audience without the need to start over. A product explainer built for one industry can be reshaped for another. A training video can be produced in multiple languages from the same source material, with the same presenter and the same structure, so a global team isn’t rebuilding the same content market by market. Update a slide, a stat, or a policy, and you’re regenerating a single scene, not re-shooting an entire video. 

What happens after someone finishes watching? 

A produced Avatar Video is a complete, useful thing on its own. It’s also a starting point. The same avatar that narrates a video can, when a team wants it, become much more. What we at Kaltura call Agentic Avatars are conversational avatars that viewer can talk to: asking follow-up questions, going deeper on a related topic, or getting quizzed on what they just watched, all grounded in the same material the video came from. A video made today can become a live conversation tomorrow. 

These are multimodal agents that can listen, speak, see and respond in context, in real time, grounded in the same organizational knowledge the video was built from. A viewer isn’t limited to asking what the video already said. They can bring a new question, a related scenario, or a task they need help completing, and the avatar works from the same validated knowledge base to guide them through it.  

These Agentic Avatars become even more powerful when they are configured to use tools. A learner can be walked through a scenario and quizzed on comprehension before moving on. A prospect can be qualified in the moment, with the avatar capturing contact details and surfacing the next relevant piece of content. A new hire can ask an onboarding avatar to walk them through a specific system step by step. In each case, the avatar isn’t just responding, it’s moving the person toward an outcome the organization cares about: a completed module, a qualified lead, a resolved ticket, a task finished correctly the first time. 

That goal-directed behavior is also where the productivity case gets real. Because the same avatar can operate across time zones and repeat the same explanation a thousand times without losing patience or consistency, it takes on the layer of communication that quietly drains organizations. AI agents handling that kind of front-line interaction have been shown to reduce average handling time by 20 to 40 percent and lower associated labor costs by as much as 30 percent, freeing people for the judgment calls and creative work that shouldn’t be automated in the first place. 

And because it’s the same agent that narrated the video, none of this requires a second setup, a second knowledge base, or a second team to manage. The presenter a viewer just watched is the same presence that can converse with them in over 50 languages, adjust to their pace, and hand them off to a person only when the moment genuinely calls for one. 

What’s the real return on scaling one person’s knowledge? 

The way I’d sum up what we built: this is a tool for turning one person’s knowledge into something that reaches far more people than that one person ever could. A trainer becomes a training program. A single expert’s explanation becomes the answer a thousand people get, consistently, in their own language, without depending on that person’s calendar. A product leader can explain a new release once and have that explanation show up as a sales enablement video, a customer onboarding asset, and an internal training module, each tailored to the audience that needs it. That’s the real value here: good communicators are rare, and everyone’s time is limited, and this is a way to stretch both further than they’d ever otherwise go.

None of that would work without the thing we started with. A hundred versions of a video, a conversation that remembers what you asked, an explanation that reaches someone in their own language, all of it still comes down to a face someone is willing to watch and listen to. 

Frequently asked questions 

Do I need any video editing experience to use Avatar Video?
No. Avatar Video is built around the assumption that the person creating the video is a subject matter expert, not an editor. You choose a starting point, a script, a presentation, a recording, a document, or a topic, and the platform handles pacing, framing, visual pairing, and B-roll selection.  

What can I feed into the tool to generate a video?
Several source types: a written script, an existing presentation, a recorded session (like a webinar or a training call), or a topic paired with your own supporting material, whether that’s other videos, documents, or a webpage. The platform reads that source content and builds a structured, narrated video from it rather than asking you to start from a blank timeline. 

Can I use my own likeness as the avatar, or only a generic one?
Both are supported. Some teams want their own trainer, executive, or product expert represented on screen; others prefer a professional presenter avatar so the content isn’t tied to one person’s calendar or likeness. Either way, the same voice, presenter, and structure can be reused consistently across every video that person or brand produces. 

How does Avatar Video handle multiple languages and regions?
Because the presenter, voice, and script are generated rather than filmed, the same source video can be regenerated in a different language with the same presenter and structure intact. That means a global team can localize a training module or a product update without re-shooting anything, and without every regional office producing its own version from scratch. 

What’s the difference between Avatar Video and an Agentic Avatar?
Avatar Video produces a finished, narrated video, the kind of asset you’d publish, embed, or send. An Agentic Avatar is what that same presenter can become when a team wants to go further: a live, conversational presence grounded in the same knowledge base, able to answer follow-up questions, walk someone through a task, or qualify a prospect in real time. Every Agentic Avatar starts as an extension of a video that already exists; it’s a downstream option, not a separate product to stand up.