DeepBrain AI starts KRW 1.5B project for conversational video production
The AMW Read
The project incrementally extends an established video platform into conversational production workflows, with impact limited by its development stage and absence of demonstrated results.
DeepBrain AI starts KRW 1.5B project for conversational video production
DeepBrain AI announced on October 19 that it is developing an AI video agent through a KRW 1.5 billion research project, including KRW 1 billion in government support. The company is the sole project operator, with work running from September 2026 through August 2028. The planned system combines a video foundation model, improved avatar synthesis and lip-sync, and agents coordinating scripts, scenes, subtitles, background music and editing. It targets marketing, education, commerce and news production.
The market significance lies in extending generative video into a complete enterprise production workflow. DeepBrain AI plans to retrieve company materials such as logos, brand colors and product information through retrieval-augmented generation, then let users request revisions through text conversations. That approach puts brand consistency, editing control and coordination between production steps alongside visual generation quality as competitive factors. Automatic recovery from production errors is a stated development goal; the announcement provides no performance results demonstrating it.
For enterprise buyers, the concrete evaluation point is the integrated beta planned for 2027, when DeepBrain AI expects to connect the agent to AI Studios and begin domain-specific trials. Buyers should test whether conversational revisions preserve approved brand assets and reduce manual editing across successive changes. The company plans integrated trials and formal testing in 2028 before global release as a standard AI Studios feature. The KRW 1.5 billion is the total research budget, rather than a disclosed equity financing round or evidence of commercial adoption.
