Iyuno Details the Architecture Behind CLOE, a Multi-Agent AI System for Consistent Content Workflows

Company outlines a system designed to maintain continuity across scenes, episodes, and languages

Iyuno 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,” said Iyuno CEO, David Lee. “We approached this as a system design problem, not just a model problem.”

Marketing Technology News: MarTech Interview with Mark Listes, CEO @ Pendulum Intelligence

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.

Marketing Technology News: How MarTech Is Enabling Autonomous Brand Engagement Across Channels?

“One of the biggest challenges in global content is that every version starts from zero—each team has to relearn the same story,” said Lee. “CLOE allows that understanding to persist, so every version starts with context instead of rebuilding it.”

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.