A Belarusian teenager who once suffered from school bullying due to unclear pronunciation has now won numerous startup competitions with an AI speech training app. This is not the plot of an inspirational movie, but a real story that happened at the Estonian startup Vocal Image.
Vocal Image has just completed a $3.6 million seed funding round, led by French edtech venture capital firm Educapital, with participation from Estonian Specialist VC and German Generations Fund. The company's mission is simple yet profound: using AI technology to help people around the world improve their voices and communication skills.
The personal experience of the company's CEO, Nick Rahoika, is the best endorsement for this product. Born in Belarus, he only started learning English after moving to Estonia and struggled with speech anxiety. However, this once introverted teenager later won many startup competitions representing his voice training startup.
Rahoika recalled during an interview with TechCrunch, "In school, I was bullied because of my unclear pronunciation." In his early 20s, as a young entrepreneur lacking confidence, he met a vocal coach named Marina Shukiyeva, who told him that voice and communication skills could be improved through training.
To help more people with similar issues, they created a YouTube channel, which eventually developed into Vocal Image. This subscription-based app positions itself as an affordable alternative to one-on-one vocal coaching, allowing users to practice safely at home. Rahoika said, "You can make strange movements and make strange sounds, but still feel safe."
Vocal Image has an interactive resource library containing tongue twisters, breathing exercises, and gesture suggestions. More excitingly, under the leadership of co-founder and CTO Mikael Kalarju, the app increasingly uses AI technology to provide automated feedback and personalized recommendations.
These guided trainings mainly focus on work-related goals, such as improving professional skills, leadership skills, and developing public speaking or presentation abilities. However, Vocal Image also supports users who simply want to boost their confidence, as well as the LGBTQ+ community. Shukiyeva had previously advocated for the rights of this group in Belarus.
Although the founding team comes from Belarus, like many other Belarusian entrepreneurs, they left their country after protests failed to remove President Alexander Lukashenko and faced brutal repression. Rahoika chose Estonia, attracted by its business environment, and it turned out to be a wise choice for a startup.
Soon after moving to Tallinn, Vocal Image joined the local incubator Startup Wise Guys, which considers the company one of its "success stories" due to its rapid growth. According to Rahoika, the startup achieved $6.5 million in annual recurring revenue with less than $1 million in pre-seed funding.
As of August this year, the startup claims to have achieved $12 million in annual recurring revenue with about 50,000 paying users. With a team of 20 people, mostly Belarusian exiles, Vocal Image now plans to expand its development team and deploy more localized versions, in addition to the existing English, Spanish, German, French, Ukrainian, and Russian languages.
This funding round comes at an opportune time, not only because the startup was recently selected as one of the five winners of Hugging Face, Meta, and Scaleway's European AI startup program, but also because it is facing increasing competition. For example, edtech company Headway recently added AI-driven speech training features to its social skills app Skillsta. However, Vocal Image can rely on its AI data repository that complies with GDPR regulations.
With about 35,000 recordings per day, Vocal Image has accumulated over 1 million real voice samples. Even better, these recordings are tagged by the community through the Voice Rating feature, a collaborative function that allows users to judge whether others' voices sound "confident" or "childish."
This is exactly the type of dataset that apps like Vocal Image desperately need to improve their accuracy. It can also help AI startups fine-tune their synthetic voices, creating further opportunities for the company beyond its B2C foundation.