# Kaon AI Describes Its Personalized Story Engine After Series B > Source: > Published: 2026-07-30 08:01:23+00:00 # Kaon AI Describes Its Personalized Story Engine After Series B Kaon AI CEO Jay Dang said in a July 30 interview that the company generates individualized story worlds in real time instead of recommending from a fixed catalog. The comments followed Kaon’s July 8 Series B announcement; Kaon says its Emochi app has more than 2 million daily users averaging 150 minutes per day and runs on a custom stack managing over 1,000 GPUs. Those operating metrics are company-reported, not independently audited. Kaon AI CEO Jay Dang described the company’s personalized entertainment strategy in a July 30 interview, expanding on a Series B announcement made three weeks earlier. The company’s central claim is that its software does not merely recommend an existing video or story. It generates a changing story world around each user’s actions. ### From recommendation to generation Kaon says its Emochi product creates character-driven narratives, images and video as people interact with it. That makes the user’s choices part of the content-generation process rather than signals used only to rank a fixed catalog. The distinction matters because the system must decide both what to generate and how to serve it quickly enough for an interactive session. In the July 30 interview, Dang said Kaon uses behavioral signals to choose the next part of a story and aims eventually to generate content that surprises users without requiring them to specify every detail in a prompt. ### The scale claims come from Kaon Kaon announced the Series B on July 8. SiliconANGLE reported that the company had raised $60 million across recent funding rounds and identified B Capital, Redpoint Ace, Goodwater Capital and DCM among its backers. The company says Emochi has more than 2 million daily active users who average 150 minutes per day. It also says its custom inference stack manages more than 1,000 GPUs across infrastructure partners, processes token volumes in the trillions each day and operates at roughly one-tenth the cost of a legacy cloud setup. These figures were repeated in independent coverage, but they originate with Kaon and have not been independently audited. Dang told PYMNTS that owning the infrastructure lets Kaon offer unlimited generation without a subscription, using advertising to support free access. That business-model claim also comes from the company. ### What practitioners should watch Kaon’s approach joins personalization, generation and serving economics in one feedback loop. For data and AI teams, the interesting problem is not only model output quality: it is whether behavioral signals can improve narrative choices without creating runaway inference costs or opaque engagement optimization. The reported session time suggests strong engagement, but it is not enough on its own to establish retention quality, user well-being or durable economics. Those questions require cohort, cost and outcome data that Kaon has not publicly disclosed in the retrieved sources. ## Key Points - 1Kaon says Emochi generates stories, images and video in response to each user’s interactions rather than selecting from a fixed catalog. - 2The company reports more than 2 million daily active users and an average of 150 minutes of daily use; the figures are not independently audited. - 3Kaon says its custom stack manages more than 1,000 GPUs and supports an ad-funded free-access model. ## Scoring Rationale The company combines consumer generative media, high reported engagement and custom inference infrastructure, but the disclosed operating metrics remain company-reported and the event has limited independently verified business detail. ## Sources Primary source and supporting public references used for this report. Practice interview problems based on real data 1,625 SQL & Python problems across 15 industry datasets — the exact type of data you work with. [Try 250 free problems](/problems)