loading
loading
Pocket FM is betting heavily on AI to improve its content economics and lift its EBITDA margin to 15%-20% from around 5% currently, cofounder Prateek Dixit told Inc42. Dixit did not disclose a timeline for achieving the margin target. However, he said AI already powers about 90% of Pocket FM’s production, while its post-production process is entirely AI-driven. Overall, AI is involved in 80%-90% of the startup’s creative processes. The audio series platform is using AI to reduce production costs, accelerate content creation, and scale its IP across languages, geographies, and entertainment formats. The margin expansion target comes shortly after Pocket FM claimed to have crossed $500 Mn in annual recurring revenue (ARR). Founded in 2018 by Dixit, Rohan Nayak, and Nishanth KS, Pocket FM offers audio content across multiple languages and genres. Backed by Lightspeed, StepStone Group, Tencent, and Times Internet, the startup has raised $196 Mn to date. It was last valued at $750 Mn during its $103 Mn Series D funding round in 2024. The startup claims its entertainment ecosystem comprises more than 7.7 Lakh titles, 5.5 Lakh creators, and 5.5 Bn hours of playtime and engagement data. It also has more than 1.4 Mn hours of text and voice content. Inside Pocket FM’s AI Content Engine As part of its broader AI push, Pocket FM has launched Sherpa, a fiction-writing platform designed to help creators build long-form stories spanning anywhere from 10 to 1,000 episodes. The platform is currently free for all writers, and Pocket FM does not expect it to become a revenue generator in the near term. “Closed-source frontier models can write one or two episodes reasonably well, but they struggle with long-form planning across 100 or 200 episodes and with creating strong IP. That was the thought process behind launching Sherpa,” Dixit said. Unlike a single generative AI model, Sherpa uses multiple specialised agents across ideation, research, story planning, prose generation, feedback, and narrative memory. The platform is intended to maintain consistency across characters, relationships, conflicts, and storylines extending over hundreds of episodes. Dixit said Sherpa’s planning and prose models use Pocket FM’s proprietary data, while its assistant layer can draw on frontier models such as OpenAI’s GPT and Anthropic’s Claude. “We have around 2.5 Mn hours of content on the platform, which is massive and gives us proprietary engagement data. We have used this data to train a model for writing great prose,” Dixit said. Taking Pocket FM’s IP Beyond Audio The startup is now looking to expand into new geographies while taking its existing IP into additional entertainment formats, Dixit said. It plans to enter Japan and South Korea over the next six to eight months. It is also exploring markets such as the Netherlands, Belgium, Turkey, and the Scandinavian countries. Beyond audio, Pocket FM is evaluating interactive storytelling and gaming opportunities, although Dixit said no specific launches have been finalised. Pocket Saga is currently the startup’s key video bet. Through the platform, Pocket FM is adapting its existing IP into long-form, AI-generated anime content. Dixit said Pocket Saga titles could run for 50 to 100 hours, comprising two-to-three-minute episodes, with individual shows potentially spanning 1,000 to 2,000 episodes. The strategy follows Pocket FM’s decision to shut down Pocket TV, its short-form video drama product. Dixit said Pocket TV suffered from weak user retention and operated in a segment where AI had increasingly commoditised short-form video production. The cofounder declined to comment on Pocket FM’s IPO and fundraising plans. Recent media reports suggest that Pocket FM is looking to raise $100 Mn to $120 Mn at a valuation of about $2 Bn. However, Inc42 could not independently verify this.
document fetched 2026-09-19 02:58 · canonical https://inc42.com/buzz/pocket-fm-eyes-15-20-ebitda-margin-global-expansion-with-ai-led-content-push