This story is part ofCNBC Make It’sThe Moment series, where highly successful people reveal the inflection point that changed the trajectory of their lives and careers.
Victor Riparbelli was almost too early to the artificial intelligence era.
In 2017, when Riparbelli co-founded his London-based startup Synthesia — an “AI video platform for business” that helps corporate workers quickly create training videos and presentations with AI avatars and text-to-video tools — ChatGPT’s public debut was still five years away. He needed funding for his business, and when he told startup investors that AI would revolutionize video technology, they often called it “a crazy sci-fi idea” in response, he says. More than 100 of them balked, he adds.
Today, startup investors flock to AI companies, which received 61% of global venture capital funding in 2025, according to a report from the Paris-based intergovernmental Organisation for Economic Co-operation and Development. Within the enterprise sector, only 39% of companies have reported AI usage having a measurable impact on their earnings, according to a November 2025 McKinsey report. Yet businesses’ spending on generative AI tools increased to $37 billion in 2025, up from from $11.5 billion in 2024, according to a December 2025 report from Menlo Ventures.
Riparbelli, Synthesia’s CEO, declined to share his company’s annual revenue, but it has corporate clients including PepsiCo, Mondelez and Heineken, according to a company spokesperson — and was most recently valued at $4 billion, from a $200 million fundraising round announced in January.
Yet the industry’s current popularity among investors didn’t help Riparbelli back in 2017. Synthesia’s survival came down to a cold email that co-founder and current COO Steffen Tjerrild sent in 2017 to former “Shark Tank” investor judge Mark Cuban.
In his email, Tjerrild promised to “keep it short,” according to a partial screenshot reviewed by CNBC Make It, and sent Cuban a demo of the product they were building — at the time, the co-founders wanted Synthesia to “make a Hollywood film from a laptop” within 10 years, as the company’s initial tagline went. Cuban responded within five minutes, and the emails exchanged over the ensuing 14 hours led him to invest $1 million in a deal that valued Synthesia at $5 million, according to the company.
Cuban already believed that “AI was the future and that video would be a core element,” and wanted to know if the Synthesia team would be “willing to work their asses off,” the billionaire entrepreneur and investor tells CNBC Make It. That proved a prescient priority when the company’s initial business model essentially flopped, and Synthesia spent four months trying to build a new product aimed at an entirely different type of customer.
Even after that new product showed promise, Synthesia remained deliberate, Riparbelli says — refusing to rush its decision to switch business models completely, until its leaders were absolutely sure the change was the best course forward. “There’s nothing easier than coming up with 20 reasons for why you shouldn’t do something,” the CEO says.
Here’s how Riparbelli spent Cuban’s investment, why the business struggled under the bright Hollywood lights, and how Synthesia found a less glamorous audience to turn itself around.
Shifting business models takes ‘a lot of deliberating’
Synthesia used the $1 million investment to hire engineers and build a product — which aimed to dub new dialogue onto existing video and translate it into other languages — that film studios ultimately found “cool and interesting” but unnecessary, Riparbelli says.
As funding dwindled, Synthesia’s few clients kept the company “afloat,” says Riparbelli. But being early to the AI era left the company in limbo, he says: The technology simply couldn’t create Hollywood-level quality video on a scale to justify his initial vision yet, and he didn’t know how long it’d take to get there.
In other words, Synthesia needed a new direction. By 2020, the company had sent its virtually dubbed demos around to plenty of companies, and some of them started asking if Synthesia could use AI to create quick, inexpensive corporate training videos and presentations, Riparbelli says. The questions made him wonder if enterprise customers represented a more eager market than Hollywood studios, he notes.
“They had completely different quality requirements, and they were very grateful that we even wanted to talk to them, and they were very excited about what we were building,” says Riparbelli.
In March 2020, Synthesia began developing a self-serve demo of an enterprise text-to-video tool. After launching a beta version in July 2020, it brought in more than $300,000 in revenue within four months from roughly 1,000 new enterprise clients, according to the company. Synthesia finally felt the “pull” of genuine product-market fit, Riparbelli says.
Even then, Riparbelli resisted making an overnight decision to switch Synthesia’s focus, because there was still risk in giving up the revenue the company already had, he says. “There’s always a lot of deliberating,” and there was no clear-cut moment where Synthesia’s leaders all agreed to “‘drop everything we’re doing and do this instead,'” Riparbelli says.
Instead, the company slowly added more offerings to support the text-to-video tool including AI avatars, editing tools and video templates. In December 2021, it launched a more comprehensive collection of its AI video tools built into one platform, completing its official shift into an enterprise model.
Since then, Synthesia has built additional business-to-business AI tools and garnered more than 60,000 enterprise customers across the world, according to a company spokesman.
Even still, the business has growing pains. AI companies face regulatory uncertainty right now, across multiple countries, making compliance challenges hard to predict, notes Riparbelli. And as Synthesia grows, so does competition from an expanding field of enterprise AI video rivals including Los Angeles-based HeyGen and London-based Colossyan. That’s where Synthesia’s early arrival might benefit the business the most, says Riparbelli, who adds that it’s actually “fun” seeing new competitors arrive on the scene.
If you’re early to a trend, and you “survive long enough for the trend to take off … you’ll understand the trend better than most of other folks,” Riparbelli says.
Disclosure: CNBC owns the exclusive off-network cable rights to “Shark Tank.”
Want to get ahead at work? Then you need to learn how to make effective small talk. In CNBC’s new online course, How To Talk To People At Work, expert instructors share practical strategies to help you use everyday conversations to gain visibility, build meaningful relationships and accelerate your career growth. Sign up today!
About The Author
