The company, which is leading the wave of micro-dramas, or serialised short-form video, has already hired close to 200 people for the Mumbai-based unit, combining creative, technical and AI-first talent, trimming production costs and time significantly.
Sources tracking Kuku's AI efforts say the company has been quietly building a proprietary AI Storytelling Stack, drawing on years of operating experience, billions of user consumption data points and learnings from thousands of content pieces. Rather than developing standalone AI tools, Kuku is building an integrated stack spanning different parts of the storytelling and production workflow.
When contacted, Kuku declined to comment on queries.
Kuku expects its AI bet to account for 80-90 per cent of overall revenue eventually and to reach more than 100 million people within six months, with output scaling to more than 2,000 shows in the same period for India and global markets, added sources who didn't want to be named as the deliberations are private.
The AI-native unit brings together writers working on story and dialogue, filmmakers from institutions such as FTII shaping visual direction, and engineers from IITs and NITs building the systems that maintain narrative continuity across long-running series.
Sources said Kuku expects the new AI unit to cut production costs for micro-drama content to roughly 30-50 per cent of comparable production costs in China, the world's largest micro-drama market and a global exporter of the category. Kuku has earlier framed AI storytelling as an extension of India's software services advantage, now applied to entertainment.
It sees a low-cost, English-proficient talent pool as India's edge in meeting the surging global demand for personalised, high-volume video, they added.
Producing one minute of micro-drama content in China typically costs upwards of $150, while Kuku is able to produce comparable-quality content for less than $100 per minute. In other markets, where compute and human resource costs are higher, producing a similar-quality video can cost anywhere from 1.5 times to 3 times more.
The savings come from two places, these people said: AI production needs far fewer people and far less time than a traditional shoot, and Kuku's systems require significantly fewer rounds of content regeneration than competitors, cutting a cost line that is typically among the largest in AI-assisted video. As a result, a single person can produce around 20 minutes of usable finished video a day, compared with roughly a minute of usable footage per shoot day in traditional production.
 "AI jobs")