Conversational Replicas by Tavus - Build with real-time digital twins that speak, see & hear

Amazing! I know how much work went into the video and voice quality, and the fast response times. Super, super impressive. Can't wait to experiment with building video AI features.

Howdy makers! We've come a long way from text-based chatbots to voice assistants, but why stop there? Humans are visual creatures, evolved over millions of years to communicate face-to-face. Shouldn't our AI interactions reflect this natural inclination? We're excited to introduce the Tavus Conversational Video Interface (CVI). With it, product teams and developers can create immersive, real-time experiences powered by AI digital twins that have a human face, vision and rich conversational awareness. They can see, hear and respond like a human would. Try our live demo at https://www.tavus.io and let us know what you think! CVI is the world’s most powerful conversational platform because it’s: - The world’s fastest (less than one second of latency between statements & responses) - The only end-to-end solution on the market (deploy a full conversational video experience with minimal code, with ASR, Vision, TTS and more all included) - The most realistic (best-in-class propietary replica model) With our end-to-end solution you can: - Build secure, realistic digital twins or stock AI agents using our replica API - Customize the LLM, persona and context for the conversation - Stream the conversation in embeddable meeting rooms (powered by Daily) - Record, transcribe and share the conversation - Handle high traffic with production-grade scalability Wondering, what can you build? 👯 🤖 Digital Twins: Extend the presence of individuals with specialist knowledge, like execs, experts, coaches, professors, healthcare professionals, and celebrities, to overcome limitations of time, space, and knowledge. Shout out to our customer, Delphi that launched their expert cloning platform last week. 👩‍💻 🎧 AI Agents: Place intelligent AI agents with a face, a voice, warmth, and humanity, where leveraging humans is not feasible today. E.g. customer support agents, digital sales assistants, personal assistants, and technical co-pilots in any industry. It was a tough technical challenge to get here and the team delivered. Huge shout out to the makers @bojoman @roey_tavus @nikita_dfp @ari_korin1 @vekaev @teodora_iliev who have worked tirelessly to deliver a world-class product for developers! Try our live demo at https://www.tavus.io and sign up to try our APIs for free. We're so excited to see what you build and would love your feedback along the way!

Connecting with the demo took a while, but it is an impressive technical accomplishment. Congrats Hassaan and the team. The replica started by making comments about the decor in my room and my hair, I was definitely pleased and shocked. Sometimes it's still speaking over me but I think it already shows the potential. I do wonder though, what's the # 1 use case people/startups/companies are using digital clones for?

I need this for my zoom meetings, so I can instead spend time coding 😂 ... then just send me the summary after 😅

Tavus's replicas feel like magic and live up to the promise and hype surrounding conversational AI. Until now I've been confident I could spot an AI-generated human, but this shatters all my expectations. It's just so realistic, and the response times are on par with human responses in a digital call. Tavus has come a long way -- these replicas are by far the best I've seen. Huge kudos to Hassaan and the team, excited to see the brilliant use cases this enables.

Hey Product Hunt friends! I'm Quinn- @hassaan_raza97's cofounder. We're beyond excited to introduce this product to the community. It's been a long time in the making, and has taken thousands of hours of research, engineering, and building. We couldn't be prouder of the whole team and our amazing partners🙌 We're most excited to see what y'all build with the Conversational Video APIs- this product unlocks transformational experiences, from K-12 education, to travel & entertainment, to eCommerce. And that's only the start. My excitement keeps building with each conversation (AI & real!) that we have. Let us know how we can best support you, and if you have any feedback- we're amped to build together! 🚀 If you're interested in learning more about our research, we have a great paper that gives some really solid insights about the model here: https://www.tavus.io/post/advanc...

Building this over the past three months has been an incredible experience. It’s a great opportunity, for sure! I don’t think I’ve poured this much effort and focus into something in a long time. I know why. A long time ago, a younger me dreamed of making my own robot—an AI I could talk to. Three decades later, this was my chance to bring that vision closer to reality. We’ve put a lot of grit into making CVI what it is: a fast-responding, highly capable, and very human-like experience, capable of being one half of a great conversation. However, the greatest achievement so far has been the many lessons we’ve learned in the process. With these lessons learned, we’re poised to make CVI even better. More naturalism, even faster responses, and smarter conversational intelligence are just the start. As we’ve pushed the boundaries of this technology, we’ve discovered more about how our own conversations work. We’ve imagined new techniques and possibilities, and we’re excited to bring those to you in the near future. I hope you have a great experience with CVI. Please reach out to us with feedback or questions. Thank you.

The first time I got my replica up and running in CVI was an eerie feeling. It was the anticipation of whether months of engineering effort would come to fruition--and whether I would to be wow’ed by something I spent such close time with, nitpicking every aspect of it under a microscope. A lot happens under the hood with CVI, but everything from interruptions, to vision, to doing research on conversational dynamics to better understanding what makes a conversation feel natural is wrapped up nicely and ultimately leaves you face-to-face with your digital twin. I booted up a call and stared at myself, and the moment my replica greeted me with an improvised “Merhaba” (Hi in Turkish), I knew this product was legit.

Pumped for this! Delphi is proud to be one of Tavus's first customers for conversational. Let's get this to the top of the product hunt leaderboard today!

🤖🤖🤖 LIVE DEMO is on our homepage >>>> www.tavus.io Try talking to Carter :)