Synthesia's PR avatar

Alexandru Voica, head of corporate affairs at video-generation startup Synthesia, sent TechCrunch reporter Dominic-Madori Davis a link to an interactive virtual avatar trained to answer common press questions about the company.

Synthesia's growth

Synthesia, originally based in the U.K., hit a four billion dollar valuation earlier this year and said last year it had crossed one hundred million in ARR.

What Synthesia does

Synthesia lets enterprises build interactive training videos with AI avatars and recently launched a product called Roleplay Sessions that lets employees practice sales pitches with an interactive AI avatar that responds and scores their responses.

Davis says yes

When Synthesia invited Davis to its new office space in New York and asked if she would like her own AI avatar, she agreed immediately, noting her outfit was cute that day and her hair was in place.

First journalist avatar

This is the first time Synthesia has made a digital avatar for a journalist, or for anyone outside of Voica. It is trained on Davis's story about why venture-backed startups commit more fraud than non-VC-backed startups.

How it was made

Davis entered a mini film studio inside Synthesia's office where they took numerous photos of her and captured a two-minute recording of her voice. They created a personal avatar and two interactive avatars, with and without glasses.

The technology stack

The interactive avatar is powered by a combination of voice-to-text, video, language, and text-to-voice models. Synthesia's tech stack includes its own video and voice models, though customers can choose alternatives from Cartesia, ElevenLabs, Google, or OpenAI.

Three product types

Synthesia builds a video-creation platform with classic avatars, an agentic platform called Sessions for interactive avatars in surveys or roleplay, and an API platform where customers combine Synthesia models with other tech services.

Testing the avatar

It took the Synthesia team a couple of days to make Davis's avatars. She tested the personal one with a script about fall arriving in New York. The voice was fairly accurate and did not pick up hoarseness from her audio sample.

Friends' reactions

Davis showed the avatars to non-tech friends who found them both interesting and creepy. Her mom called it amazing and joked after testing it: I don't remember giving birth to two of you.

“I don't remember giving birth to two of you.”

Questions about journalism

Davis wonders if people would accept avatars presenting the news and whether CEOs would prefer talking to an AI avatar of a journalist rather than a human one. One investor said no immediately.

Trust cannot be outsourced

Davis values connecting with people, writing stories, and researching new topics. But the biggest part of journalism is trust, which does not seem like it could ever be outsourced to an AI.

Gen Z and digital twins

Davis predicts her generation, Gen Z, likely won't get used to digital twins because they feel sci-fi. She finds digital avatars less jarring than humanoids, and notes that if stuff gets weird, she can always log off.