Anton Volovyk
Co-CEO of Reface
Where to hear and read about Anton Volovyk1 episode or test
Anton Volovyk is co-CEO of Reface, a consumer app company that grew on its own face swap model in 2020 and now runs apps like Bite Pal. The most useful thing to learn from him is to compete on product, distribution and small operational gains, with AI as the machine around a human core idea.
No one actually remembers what you clicked on.
The AIGC Ad Creatives Crushing It vs. The Ones Flopping
Anton Volovyk: the story and the lessons
1How Reface learned it is a consumer product company
Reface grew big in 2020 on its own face swap model. The early success came from a model the company built itself, and it shaped how the team saw its own edge.
Co-guest Nathan Hudson says that 2020 win fed an ego problem later on.
Volovyk's team wrongly believed it had to keep building its own models. As the industry changed, Reface kept thinking it needed its own ML research. It took time to see that it is a consumer product company.
The edge lies in packaging, UI and UX, and the distribution machine, not in which AI model is used.
Volovyk says a company born today would be much smaller. He says a new company could produce the same output with far fewer people, while Reface carries processes and legacy that startups don't.
Hudson calls this team-size shift the big bet he is making for next year.
He sees technology running well ahead of how apps use it. Volovyk says few examples exist of agents or assistants meaningfully improving an app. Even if progress stopped now, he sees five more years of innovation to apply.
2How Volovyk backed Bite Pal and what he thinks wins
Reface funded Bite Pal for a year before its economics turned positive. Ads ran the whole year, paid for by other apps that generated cash. For a long time Bite Pal was a heavily AI-negative product.
Volovyk admits sunk cost helped, since the team had invested too much to step back.
He warns against launching something generic and finding fit through ads. Many founders ship a generic app, look for fit with ad spend, and only then build the product. Many never reach fit.
Bite Pal later reached number three among nutrition apps and top 10 among fitness apps globally by downloads.
The core concept stays human, and AI works around it. Bite Pal's raccoon, a Tamagotchi-like pet you care for, was a human idea. AI frames everything around it but cannot find it.
The team found that relying fully on AI for ideas did not work. In their words, AI is a multiplier of your own talent.
Volovyk calls distribution the lasting moat now that building is cheap. Anyone can build a Bite Pal-type app in a few days with a few developers, so competition moves to marketing. He says the marketing team's role expanded and became the key competitive moat.
Easier coding does not mean you can crack distribution.
He compounds small operational gains, citing the book 7 Powers. Being about 2% better at every step adds up, from a slightly better interface for TikTok ads to slightly better creative performance.
He says that extra margin let Bite Pal scale 10 to 100x.
3How Reface runs its AI marketing machine
Volovyk names n8n as the highest-impact AI tool. He calls it a factory linking Reface's models and automations from creatives through user acquisition to monetization. No off-the-shelf marketing tool did this, so a horizontal automation tool won.
Reface plans to move to something more robust and durable later.
He ranks the funnel layers and puts AI at the outer ones. His order is human need, product, funnels, then creatives. AI is powerful at the outer layers, and the inner ones stay with people for now.
He expects the conversation to look different in two or three months.
Reface wins on creative diversity from an in-house, many-model tool. A designer drags and prompts across models such as Kling, ElevenLabs, Suno and Nano Banana, with an asset library in the same tool. Reface makes fewer creatives than big publishers but more varied ones.
Sound and music drive performance more than expected, so many tracks get tested, including AI-generated songs.
Volovyk builds brand after the click, not through creative rules. Strict brand guidelines narrowed Reface's creatives. Performance marketing began working once most constraints were dropped.
Bite Pal still gets a lot of organic and word-of-mouth traffic. His view: "No one actually remembers what you clicked on."
He found that disclosing AI use up front does not hurt conversion. Reface apps show popups saying Gemini may send third-party data to Google. A/B tests showed no big drop-off.
He compares it to GDPR blindness: people accept the notice and move on.
4How Volovyk hires and thinks about customers
Volovyk hires for curiosity and drive over AI skills. People from unrelated jobs joined Reface, showed energy and learned fast. The company supplies the tools, experts and resources.
He says ads work when they reflect how the viewer lives. A keen cyclist himself, Volovyk can tell when an advertiser understands his life. He says winning creatives need that real understanding, such as of someone who fasts or cycles.
This is the human part of ad making that the team has not been able to hand to AI.
The AIGC Ad Creatives Crushing It vs. The Ones Flopping
Two growth operators show real AI-generated ad creatives and explain how they build, test and scale them, from UGC anchors with swapped backgrounds to fully prompted videos. The takeaway is that creative production is cheap, winners cannot be predicted, and the edge comes from concept, volume and distribution.
Reface grew big in 2020 on its own face swap model. The early success came from a model the company built itself, and it shaped how the team saw its own edge.
Co-guest Nathan Hudson says that 2020 win fed an ego problem later on.
Volovyk's team wrongly believed it had to keep building its own models. As the industry changed, Reface kept thinking it needed its own ML research. It took time to see that it is a consumer product company.
The edge lies in packaging, UI and UX, and the distribution machine, not in which AI model is used.
Volovyk says a company born today would be much smaller. He says a new company could produce the same output with far fewer people, while Reface carries processes and legacy that startups don't.
Hudson calls this team-size shift the big bet he is making for next year.
He sees technology running well ahead of how apps use it. Volovyk says few examples exist of agents or assistants meaningfully improving an app. Even if progress stopped now, he sees five more years of innovation to apply.
Reface funded Bite Pal for a year before its economics turned positive. Ads ran the whole year, paid for by other apps that generated cash. For a long time Bite Pal was a heavily AI-negative product.
Volovyk admits sunk cost helped, since the team had invested too much to step back.
He warns against launching something generic and finding fit through ads. Many founders ship a generic app, look for fit with ad spend, and only then build the product. Many never reach fit.
Bite Pal later reached number three among nutrition apps and top 10 among fitness apps globally by downloads.
The core concept stays human, and AI works around it. Bite Pal's raccoon, a Tamagotchi-like pet you care for, was a human idea. AI frames everything around it but cannot find it.
The team found that relying fully on AI for ideas did not work. In their words, AI is a multiplier of your own talent.
Volovyk calls distribution the lasting moat now that building is cheap. Anyone can build a Bite Pal-type app in a few days with a few developers, so competition moves to marketing. He says the marketing team's role expanded and became the key competitive moat.
Easier coding does not mean you can crack distribution.
He compounds small operational gains, citing the book 7 Powers. Being about 2% better at every step adds up, from a slightly better interface for TikTok ads to slightly better creative performance.
He says that extra margin let Bite Pal scale 10 to 100x.
Volovyk names n8n as the highest-impact AI tool. He calls it a factory linking Reface's models and automations from creatives through user acquisition to monetization. No off-the-shelf marketing tool did this, so a horizontal automation tool won.
Reface plans to move to something more robust and durable later.
He ranks the funnel layers and puts AI at the outer ones. His order is human need, product, funnels, then creatives. AI is powerful at the outer layers, and the inner ones stay with people for now.
He expects the conversation to look different in two or three months.
Reface wins on creative diversity from an in-house, many-model tool. A designer drags and prompts across models such as Kling, ElevenLabs, Suno and Nano Banana, with an asset library in the same tool. Reface makes fewer creatives than big publishers but more varied ones.
Sound and music drive performance more than expected, so many tracks get tested, including AI-generated songs.
Volovyk builds brand after the click, not through creative rules. Strict brand guidelines narrowed Reface's creatives. Performance marketing began working once most constraints were dropped.
Bite Pal still gets a lot of organic and word-of-mouth traffic. His view: "No one actually remembers what you clicked on."
He found that disclosing AI use up front does not hurt conversion. Reface apps show popups saying Gemini may send third-party data to Google. A/B tests showed no big drop-off.
He compares it to GDPR blindness: people accept the notice and move on.
Volovyk hires for curiosity and drive over AI skills. People from unrelated jobs joined Reface, showed energy and learned fast. The company supplies the tools, experts and resources.
He says ads work when they reflect how the viewer lives. A keen cyclist himself, Volovyk can tell when an advertiser understands his life. He says winning creatives need that real understanding, such as of someone who fasts or cycles.
This is the human part of ad making that the team has not been able to hand to AI.