.svg){kind=icon}
AI Avatar Masterclass: How Digital Humans Come to Life
Loading video...
Show Notes
What does it take to give AI a face, and build a digital character people actually want to talk to? In this special montage episode of the Convo AI World Podcast, we bring together experts from Trulience, Akool, and AKA Virtual to unpack the technology, design choices, and product thinking behind the rapidly evolving world of AI avatars and digital humans. Jeff Lu of Akool explains why his team has focused on photorealistic people, real-time video generation, and running increasingly capable AI experiences closer to the edge. Jia Shen of AKA Virtual makes the case for a very different approach: characters do not always need to imitate humans. He explains why caricature, animation, and clearly defined character roles can sometimes create more believable experiences. Richard Bowdler of Trulience explores another part of the stack: giving conversational AI a face and body while keeping the underlying intelligence modular. That means combining avatars with LLMs, speech technologies, knowledge, and other components while allowing developers to choose the "brain" behind the character. Whether you're building conversational AI agents, virtual assistants, interactive entertainment, customer experiences, or the next generation of AI characters, this masterclass offers a practical look at what it takes to turn an AI model into a digital presence people can actually interact with.
Key Topics Covered
- •The AI avatar stack: LLMs, speech, animation, video generation, and real-time interaction
- •Building photorealistic people and real-time visual experiences with Akool
- •Why AKA Virtual favors anime-inspired characters and caricature over digital clones
- •Trulience's modular 'bring your own brain' approach to conversational avatars
- •Connecting avatars to LLMs, voice AI, RAG, and other conversational systems
- •Designing specialized AI characters for specific roles and experiences
- •Applications across customer experiences, entertainment, gaming, virtual assistants, and interactive services
Resources & Links
Episode Chapters & Transcript
Teaser: What gives an AI avatar life?
Jia Shen, Jeff Lu, and Richard Bowdler on giving avatars intelligence, building the underlying tech stack, and creating characters people actually want to engage with.
Welcome: An AI Avatar Masterclass
Hermes Frangoudis introduces a special montage exploring how builders are giving conversational AI a face, voice, and more human way to interact.
From generative video to real-time avatars (Akool)
Jeff Lu traces Akool's origins in generative video and its push toward photorealistic characters, real-time interaction, and AI models running on edge devices.
Why cartoons can beat photorealism (AKA Virtual)
Jia Shen makes the case for stylized characters, clearer expectation setting, and why caricature can create stronger experiences than digital clones.
Giving AI a face and a body (Trulience)
Richard Bowdler shares Trulience's founding story, from an idea inspired by elderly care to a broader vision for making AI interaction more human.
Building the models behind AI avatars (Akool)
Jeff Lu explains Akool's mix of in-house technology and open-source foundation models, including its decision to develop the core avatar stack internally.
Virtual characters and the changing meaning of real (AKA Virtual)
Jia Shen explores virtual economies, younger audiences' relationship with digital experiences, and how generative AI could reshape perceptions of authenticity and value.
How avatar technology evolved (Trulience)
Richard Bowdler traces avatar development from camera cages and motion capture to CGI and today's more accessible, conversational AI-powered approaches.
Quality, speed, cost, and controllability (Akool)
Jeff Lu breaks down the tradeoffs between visual quality, generation speed, compute cost, ease of use, and the control professional users need.
Giving virtual characters intelligence and purpose (AKA Virtual)
Jia Shen explains AKA Virtual's DI platform, functional vs. entertainment characters, multilingual experiences, and how different AI models can work together behind a character.
Bring your own brain: Building the real-time avatar stack (Trulience)
Richard Bowdler explains Trulience's modular approach to LLMs and voice AI, along with the latency and lip-sync challenges behind natural avatar conversations.
Real-time avatars and video translation (Akool)
Jeff Lu explains the shared foundations behind streaming avatars, AI agents, live translation, voice cloning, and whole-face reanimation.
Building multilingual avatar conversations (Trulience)
Richard Bowdler explains how Trulience works with external language and voice providers to let avatars listen and respond across languages.
Why character voices need specialization (AKA Virtual)
Jia Shen explores AI singing and speech models, and why anime characters, celebrities, and entertainment voices demand more specialized voice design.
Teaching avatars emotion and expression (Trulience)
Richard Bowdler breaks down image- and video-based avatars and the challenge of synchronizing facial expressions with the right conversational emotion.
Where AI avatars are being used today (Akool)
Jeff Lu maps avatar adoption across marketing, advertising, film production, AI agents, internal communications, and multilingual video.
Making knowledge accessible through digital humans (Trulience)
Richard Bowdler shares an India healthcare deployment and explores how natural-language avatars can open access to information beyond text-based interfaces.
Scaling digital humans with client-side rendering (Trulience)
Richard Bowdler explains how rendering avatars in the browser can reduce infrastructure requirements and help interactive digital humans scale.
How Akool turns experiments into products (Akool)
Jeff Lu explains how Akool tests new capabilities, gathers user feedback, and moves promising experiments from beta into its core platform.
Why would anyone talk to an AI character? (AKA Virtual)
Jia Shen uses fortune tellers and entertainment characters to show why successful AI characters need a clear role, interaction model, and meaningful payoff.
Where digital humans can have the most impact (Trulience)
Richard Bowdler explores customer service, healthcare, e-commerce, onboarding, learning, and the potential for AI avatars to democratize access to expertise.
Click on any chapter to view its transcript content • Download full transcript
Convo AI Newsletter
Subscribe to stay up to date on what's happening in conversational and voice AI.