How AI Digital Twins Became a Multimillion Dollar Business – Interview with David Riha
- 1 day ago
- 6 min read
What happens when expertise stops being delivered through courses and starts living inside an AI that talks, remembers, and grows with each user? In this conversation, the creator behind Buddy and early AI digital twins explains how stand-alone “AI twin” products quietly turned into a multimillion-dollar subscription category in just months. From breakthrough launches to the shift toward relationship-based AI coaching, the discussion reveals why the next wave of digital products may not look like products at all.
David Riha, AI Digital Twin Builder
What made you realize AI digital twins could become a business category of their own rather than just another AI tool?
That's quite a story! The first AI digital twin that I've built was for my brother, he's a famous expert on the European market, teaching people how to build successful online businesses. It was his idea, but he was not sure if people would actually buy it, this was February 2025, and nobody had done this before. At first, he was thinking he would offer it as a bonus with some online course, but then he thought, what if I actually just sell it as a stand alone digital product and prove that it's possible? It was a scary idea, but he went for it, and it turned out to be the revolution. He made $180K in subscription revenue within the first 30 days.
Then, 3 months later, we launched another one for another big expert, also our friend, and she made $258K in 20 days. Just from selling her AI twin, stand alone product, annual subscriptions for $1,190. Same strategy, even better results.
Today, more than 150 experts and coaches launched their AI twins, many of them selling it as a stand alone product for $1 to 2K, annual subscriptions, and they made more than $4.6M in subscription revenue.
After launching more than 130 AI twins, what separates the most successful ones from the rest?
Great question. The most successful ones accepted that this is possible and decided to put focus on launching it big, instead of hesitating, forever improving, delaying it, and offering it to a couple of people here or there. That's what matters the most if you want to have big success with this, launch it big, go all in.
Don't hesitate, don't be afraid that nobody will buy it. So many people actually want your AI once they understand that it can deliver the transformation they've been hoping to get from you. Sometimes it does this even better, because it fills the gap between learning and implementing. It's there in their pocket, guiding them when they need it the most, when they actually need help with their specific situation.
You often describe Buddy as an AI companion rather than a chatbot, so what difference does that create for the user experience?
The difference is that all the mainstream AI chats like ChatGPT or Claude talk like AI. When it comes to conversation, it's all just Q&A chatbots.
We've built Buddy to chat like people do, with all that that means. Remembering everything about the person, and always using it in the conversation, no conversation ever starts from scratch. Not saying "I understand" to every struggle the user talks about. Being empathetic, but not fake. Knowing how to push back and not always just assure the user they're right. Sometimes writing short answers, sometimes long. Sometimes sending follow up messages. Sometimes initiating the conversation hours or days later. Never sharing what the user says with anybody, keeping conversations super private.
The result is that people form actual mentoring relationships with Buddy. They keep sharing more and more of their life story with Buddy, and they go really deep on personal topics. The AI knows all about them and adjusts the conversation, advice, and coaching to the situation. This creates ultra personalized coaching that's so spot on and connects the dots so well that it delivers real transformation very effectively.
Now, on top of that, the experts form it into their own version of Buddy, with their expertise, their personality, their approaches, and their coaching. The result is an experience that's very similar to talking to the real expert, and it gives people very similar value.
How do you see relationship based AI coaching changing the coaching and consulting industry over the next few years?
Many coaches and experts will have their AI twin. What this is doing to the coaching industry is actually good for everybody who does this, people get access to world class coaching from the world's best experts at a fraction of the cost with 24/7 availability, and the coach can scale their business beyond anything imaginable, without having to put continuous delivery effort into it. The coach could technically retire if that's what they'd want to do, but have even more profitable business than ever before.
For coaches that decide not to build their AI twin, I believe this will still work. AI twins are just one of the ways, and if the coach enjoys one on one coaching, there's no reason to stop doing that.
What do most experts misunderstand about turning their knowledge into a scalable AI product?
Most experts don't understand that the important thing is the quality of the AI twin. You could probably build this over a weekend from scratch using Claude Code or use some of the mainstream tools like Delphi. But making it so good that people pay $2K a year for it is a completely different game.
Most experts use some basic tool to build a Q&A chatbot from their know how, and think AI twin is a mediocre product that can be a little bonus in their membership section.
I've heard this so many times, "Yeah, I already have that, it's integrated in my membership section." But it's not good. People ask a question once in a while, and that's where it ends. It's not a digital product, it's a feature or a tool.
An AI twin that justifies $2K pricing needs to be in a messaging app, remember everything about the user, and use it. It needs to form a relationship, initiate conversations, and talk like a human, not like a chatbot. It needs to deliver real transformation, not just answers to questions.
For a coach or author considering an AI twin, where should they start before thinking about the technology?
Think about your audience and calculate your potential revenue. How big is your audience? How many people in your audience are buying your digital products? It sells better than other digital products like online courses, so if you've been previously selling that, it's going to work. If you previously sold an online course to 500 people, you will likely sell 500 or more subscriptions to your AI twin. If it's built well, then don't hesitate to go all in and make a commitment, you will not play a small game with this.
What have you learned from building AI versions of experts like Jen Gottlieb and Marie Diamond?
There is still a huge untapped potential, and the big world famous experts I'm working with are still in the process of launching it soon. Those people who made multiple six figure revenues, many of them are not world famous coaches. They are smaller experts in a niche or a specific, smaller market. That's just how it works, the early adopters in any industry are always the smaller players, because they can take risks. Then it becomes big once the bigger players see it works, and it's not risky business anymore. That's exactly what we are seeing now.
We have one world famous expert on the platform (I can't name them yet) who launched their AI twin privately to their existing clients, and now they're close to $1M in revenue. Before public launch! Once a world famous expert with a huge audience does a public launch (we have it planned with multiple of them for 2026), their revenue will be in millions or tens of millions of dollars.
How do you balance innovation with the responsibility that comes from creating AI systems people may rely on for guidance?
That's a great question. We could speed up development of the BuddyPro platform by a lot with higher use of AI development, but we want to ensure that everything remains secure, information stays private, and we develop new features and adjust behavior responsibly. We are not using AI agents to develop the platform as much as we could, because we want to ensure the highest quality on the market and ensure that we keep providing a secure and responsible solution.
When it comes to acting on the guidance of the AI, the user must still be aware that they're talking to an AI and make their own decisions. People should not blindly follow advice, not from a real expert and not from an AI. We optimize BuddyPro to never say wrong things that could harm somebody, but still, humans should use critical thinking when talking to AI.
What is one shift in mindset that experts need to make to thrive in the age of AI?
They need to accept that it's here and they should start thinking about how they can use it, instead of trying to ignore it. Yes, AI might replace you to some extent, but if that's going to happen, make sure you own the AI that's replacing you. Then it's not too bad, people talk to your AI instead of you, you get paid, and you only work if you really want to.
Read more from David Riha
