Iyuno today shared new details on the architecture behind CLOE, its AI platform designed to address a key challenge in content production: maintaining consistency at scale.
While AI tools can generate outputs quickly, they often struggle to preserve continuity across longer-form content—leading to inconsistencies in tone, character, and narrative across workflows.
"At scale, content breaks when understanding resets. We approached this as a system design problem, not just a model problem."
— David Lee, CEO, Iyuno
CLOE is built on a multi-agent framework that structures how content is interpreted before outputs are created. The system operates across three core layers:
Sensory layer: captures inputs across dialogue, visuals, and audio
Fusion layer: connects those inputs, tracking relationships and narrative flow
Memory layer: stores this understanding as a persistent, reusable knowledge base
This approach enables CLOE to maintain a continuous representation of content, allowing multiple workflows—such as subtitling, dubbing, and accessibility—to draw from the same underlying context.
"One of the biggest challenges in global content is that every version starts from zero—each team has to relearn the same story. CLOE allows that understanding to persist, so every version starts with context instead of rebuilding it."
— David Lee, CEO, Iyuno
CLOE supports a growing set of modular AI capabilities, or “skills,” across the content lifecycle, with each operating from the same shared foundation.
Iyuno is continuing to expand the platform’s capabilities as it scales deployment across global content workflows.
