Anam raised $9M to build AI avatars that respond in real time using just a single photo. The startup uses custom diffusion models to create expressive avatars whose facial expressions sync with conversation, differentiating itself from competitors like HeyGen and Synthesia by focusing on dynamic interaction instead of pre-recorded video.
True or false, you came up with this idea at a pub?
True. We love it. 7 drinks in.
Can you tell us a little bit about what Anim does?
We're building human faces for your product. We leverage our own custom diffusion models to build AI avatars who feel as natural, or if not more natural, than engaging with a human.
I guess I'm curious, what was the moment you knew this needed to exist?
I don't know if I want to watch a video unless it's like a 7-second TikTok video. I want to interact. I want to talk to someone. I want to feel what the brand feels like. It was so clear that there was nobody doing this properly, so we decided to take it on ourselves.
I feel like every time we talk, there's like some new tech breakthrough, which is amazing. It's obviously been a very busy start to the year, so tell us a little bit about what you guys have been up to.
The one-shot model is what we're building towards. You upload one picture, you can build a real-time AI persona from that picture. Because of our model, every single pixel is generated in real time. It means that you'll get an output of an AI persona who is fully expressive and emotive. They can move in conjunction with the topic of conversation. Most companies we speak to speak to, they use large language models and it's just like text. But actually, if you want to build something that feels like your product, you should be able to go in, develop it yourself, add how it looks, how it feels, how to interact with it, and then you can just deploy it in your own product.
It's obviously a really exciting space, but also like a decent amount of competition in the category. I'm curious, what's different about Anna?
You've got incredible companies like ElevenLabs who do text-to-speech. They're adjacent to us. We also have HeyGen who sits somewhere between Synthesia and us. And so what's actually different is that we're the only company who's fully thinking about real-time interaction. To do that, we've built our own diffusion model, the whole networking layer. What does that actually mean for a customer? It means that you have these AI personas that actually have expressivity and that you can control based on the conversation. And so it's not just this video loop with mouth dubbing, it's actually engaging and interactive. To me, our competition isn't all these other great companies that exist. It's, it's real humans, right? How do they engage? How do they interact? Because that's what I want to talk to.
The product is amazing, amazing, amazing. But I know the thing you care most about is the team. Can you tell us a little bit about them?
Myself and Ben actually worked together at Synthesia. I led go-to-market. He was a research engineer. He's worked at like 3 AI labs. And so we've amassed this team of quite senior research engineers and also software engineers. One criteria we look for is that they've been a founding engineer before and also worked at FAANG. That way you get someone who's seen velocity, high pace, but also has seen scale and how structures can be built. But yeah, it was really important to build a, a great research and engineering team because that's our core product. And now it's about how do we ship that to market? We're always looking for smart, intelligent people who are capable. We have offices in London. We're opening up a New York office. What I want to build for a team is the team that I always wanted to be a part of, which is this high-performing sports team, right? It's where you're trying to build something that you're proud of, that you respect. Building out that commercial arm will be the next challenge.
Last question. Where can folks go to learn more about Anam?
So you can go to our website, which is anim.ai, and try out our demo, talk to Cara, she'll play a game with you and just, yeah, try out the, the next interface for technology. It's quite exciting.
