Synthesia’s training tools are adding live AI coaching to the mix

Synthesia’s training tools are adding live AI coaching to the mix

Synthesia is taking its AI training business in a new direction. The British startup, best known for helping companies quickly produce training videos with AI avatars, has introduced a more interactive product aimed at live practice rather than passive viewing.

Called Roleplay Sessions, the new offering lets employees rehearse situations such as sales pitches, performance reviews, and customer complaints with an AI avatar that responds in real time, challenges their answers, and then evaluates them using a rubric. According to the report, Roleplay is the first product in a larger Sessions platform, which Synthesia plans to extend to other areas such as job interviews and candidate screening.

The launch reflects a broader shift in enterprise AI, where companies are increasingly focused on whether tools actually improve performance. Synthesia says its view is that training works best when people actively practice a skill and receive feedback, rather than simply watching a video or reading material. The company argues that its new format is designed to capture that behavior change and provide measurable results.

That measurement angle also gives Synthesia a more strategic position in the market. While its avatar and voice technology are proprietary, the reasoning layer behind the system comes from OpenAI. By adding rubrics, performance data, and analytics on top, Synthesia is presenting the product less as a video generator and more as a platform for tracking and improving employee performance. The company says early customers already include major enterprises, and the most common uses are sales training and leadership development, especially around soft skills.

For now, Roleplay is being sold to enterprise customers, but Synthesia expects to broaden access in the coming months to smaller businesses, prosumers, and possibly educational institutions as inference costs decline. The company’s pitch is that the next phase of AI adoption will be judged less by flashy demonstrations and more by whether the software delivers measurable business outcomes.

Source: techcrunch.com