For years Synthesia’s pitch was simple and it worked: use AI to help enterprises create interactive training videos, in minutes, for a fraction of the cost normally associated with developing corporate educational materials. On Wednesday the British startup made a different bet. The new product isn’t another faster video maker. It’s an attempt to prove value, not just produce content.
That product is Roleplay Sessions, an interactive training product where employees can practice high-stakes conversations like sales pitches, performance reviews and customer complaints with an AI avatar that talks back, pushes back and then scores them against a rubric. Roleplay Sessions is the first release under a broader “Sessions” platform Synthesia plans to expand into other formats, including job interviews and candidate screening, TechCrunch has exclusively learned.
It’s a shift in timing as much as in product. The launch comes as enterprises are taking a harder look at whether AI spend is paying off. Synthesia itself is leaning on learning science to make the case. Citing a meta-analysis of learning research, the company argues that most corporate training — including, implicitly, its own core video product — stops short of actually changing behavior. The trainings tend to inform and demonstrate, but what really pushes someone into actually learning something is doing that thing, along with feedback.
“Video performs way better than text or sending out a document,” Synthesia CEO and co-founder Victor Riparbelli told TechCrunch. “But for most things, we learn the best by actually practicing something rather than just reading it.” That line is the whole thesis. If companies want skills to stick, they need reps, not just polished explainers.
Where video hits a ceiling
Video is cheap now. Synthesia helped make that normal. Teams can spin up avatar-led modules in minutes, brand them, ship them globally. That solved a distribution problem. It didn’t solve the transfer problem.
Riparbelli puts it plainly: watching is not doing. Enterprises found that out after the first wave of AI demos looked great on social media and then sat unused inside L&D dashboards. The moat, he argues now, isn’t generating content. It’s proving it worked. The startup’s newest product is making a different bet — that the real moat in enterprise AI isn’t just generating content, but in proving that it worked.
What Roleplay Sessions actually does
Roleplay Sessions is built around conversation, not consumption. An employee picks a scenario, the avatar plays a counterpart, and the exchange is live. The system pushes back, interrupts, asks tough questions, mirrors real friction. After the session it scores the employee against a rubric.
The scoring is the part Synthesia wants you to notice. It’s not a vibe check. It’s structured feedback tied to behaviours the company cares about: active listening, objection handling, tone in a performance review, de-escalation with an angry customer. That data can be aggregated.
“From a managerial executive layer, we’re seeing huge interest in mapping out the talent in their company in a way that’s been really difficult to do before,” Riparbelli said. “If you have your entire sales team practice with an AI role player, you can get very granular data on how your sales force is doing overall.”
The company’s avatar and voice technology are proprietary, but the underlying reasoning intelligence is OpenAI’s. Synthesia is tacking on a layer that includes rubrics, performance data and analytics. The positioning moves from AI avatar company to performance-management platform with a video front end.
Early use and customers
Synthesia already has a few early Roleplay customers with scaled commercial rollouts, including “one of the top three companies by market cap in Europe, one of the top five Fortune 100, and one of the biggest recruitment companies in the world,” per a company spokesperson.
The two most popular use cases involve training sales teams and leadership — programs like practicing how to sell a product or have a difficult conversation. Much of it is soft skills training, Riparbelli said. That makes sense. Soft skills are hard to teach at scale because they require judgment under pressure. A video can model it. A roleplay can test it.
Customers can either create their own training programs on the platform or enlist the help of Synthesia’s consultants to build a bespoke solution that relies in part on whatever training documents and context the company already has. In other words, the same configure-or-consult playbook the company used for video, now applied to interactive scenarios.
Platform ambitions beyond roleplay
Roleplay is the first product under Sessions. Synthesia plans to expand into other formats, including job interviews and candidate screening. That fits the broader narrative: sessions, not assets.
Roleplay is an enterprise offering for now, but the goal is to offer it more widely to small businesses, prosumers and potentially educational institutions in the next few months as inference costs continue to fall. The cost curve matters here. Real-time, multilingual avatar conversation is expensive. As it gets cheaper, the use cases widen.
Riparbelli believes products like Roleplay represent the next phase of enterprise AI adoption, where companies care less about whether AI looks good in a social media demo and more about proving it generates measurable business outcomes. That’s a market read as much as a product bet. After a year of pilots and press releases, buyers want numbers: completion rates aren’t enough, behavior change is.
Why this matters for Synthesia
The move also buys defensibility. Avatar and voice tech are proprietary, but they’re not uniquely so. The underlying reasoning is OpenAI’s. By adding rubrics, performance data and analytics, Synthesia is building a layer that’s harder to replicate with a generic LLM and a talking head.
It’s also a candid admission about the limits of its first act. If video performs way better than text, it still isn’t practice. Roleplay tries to close that gap. Whether it can do so at enterprise scale, with consistent quality and without hallucinating feedback, will determine if the Sessions platform becomes the core business or a niche add-on.
For now the story is clear: Synthesia wants to stop being known for making training look good and start being known for making training work. That means less focus on minutes-to-video and more focus on data that managers can actually use to map talent, coach teams and justify spend. The pitch has changed from speed and cost to proof and performance, and the company is betting the enterprise market is ready for that conversation.
