How Beyond Presence is building instant real time video avatars

Digital avatars have existed for years, yet most interactive video characters still struggle with uncanny facial movements and noticeable delay during live conversations. Munich-based startup Beyond Presence is building foundational artificial intelligence models designed to make face-to-face digital interactions feel fluid and authentic. Founders Awais Shafique and Felix Altenberger set out to address complex computer vision challenges to make conversational avatars truly responsive in real time. The team focuses on creating realistic video agents that can hold natural discussions across customer service, recruitment, sales, and online education.

The core of the technology relies on proprietary real-time avatar models engineered to deliver expressive video responses with turn latency under one hundred milliseconds. By combining advanced speech processing with generative rendering, the platform matches natural human facial expressions and precise lip movement without awkward pauses. The company recently introduced its Genesis 2.0 model, which supports over one hundred languages for global enterprise deployments. This infrastructure allows businesses to add expressive video presence to existing voice bots, enhancing overall user engagement and trust.

Integrating these avatars into modern software applications is designed to be straightforward for technical development teams. Beyond Presence offers flexible options through simple Application Programming Interfaces, allowing direct connections with audio agent frameworks such as LiveKit and PipeCat. Developers can also connect end-to-end agent webhooks directly into workflow automation tools like n8n using standard API keys and system prompts. This architecture transforms routine one-on-one virtual interactions into parallel conversational streams that scale seamlessly across enterprise systems.

The founding team brings deep expertise in computer vision and software engineering to this ambitious research project. Awais Shafique previously co-founded the 3D body scanning startup Presize, which was subsequently acquired by tech giant Meta. Building upon experience gained from that acquisition, Awais Shafique partnered with Felix Altenberger, who previously worked across engineering roles at ZenML, Presize, and Celonis while studying at the Technical University of Munich. Their shared background through the Center for Digital Technology and Management helped shape their collaborative approach to generative video.

To support its ongoing research and model training, Beyond Presence secured three point one million dollars in pre-seed funding. German venture capital firm HV Capital led the investment round, with additional participation from 10x Founders and Alba VC. The company also attracted individual angel backing from key figures at Meta, DeepMind, and Zalando, alongside former Presize co-founders. Operating from Munich, the startup places strong emphasis on regulatory compliance, building its platform in full alignment with European Union General Data Protection Regulation and EU AI Act standards.

As interest in interactive video interfaces accelerates across global markets, the startup is expanding its platform capabilities for enterprise clients. Beyond Presence will demonstrate its real-time avatar technology at Web Summit 2026, showcasing how sub-hundred-millisecond response times transform conversational software. By combining ultra-low latency rendering with multilingual support and secure enterprise architecture, the team aims to set a new benchmark for humanlike digital interaction.

Find out more at beyondpresence.ai