“I still feel that our content catalog is not sufficient for our users. There are so many genres and subgenres that we don’t have in our library. I don’t think we have a depth of content that is in the mature entertainment category,” Nayak told TechCrunch over a call.

The company already has an outstanding partnership with ElevenLabs to convert writings into audio series. This has resulted in 5x quicker production and 30x lower cost as compared to a professionally generated audio series.

“We have already tested how these AI adaptations perform in different markets, and we have seen encouraging results. We are still refining our models for errors, but we feel the technology is good enough to use in the production of shows,” Nayak said.

One of the AI tools Pocket FM is trying out caters to the adoption of stories in different regions. The company said it has trained in-house models that don’t look at mere translation but handle cultural nuances when converting stories of one region to another.

He added that it is challenging to solve the hallucination of the models in the context of stories that span hundreds of episodes. Pocket FM said it had to tackle the limitation of context windows of open source models and also build maps of relations between different entities in a story to maintain character consistency.

Another tool for writers the company is testing works as a creative assistant to them, helping them with alternative storylines or giving them ideas about plots. The company also plans to infuse some insights from its historical data in the tool to indicate to writers what works on the platform.

Nayak mentioned that while this tool is in its early stages, the company wants to give the power of a writer’s room to a solo writer, who might be putting out an episode of a day. He noted that the writer’s room gives you the ability to brainstorm without taking out creative controls, and that’s the core idea behind this writing assistant.

Additionally, the company is investing in creating a blockbuster engine that will feed on insights from the platform about what shows are becoming a hit.

The end goal of Pocket FM is to scale its catalog, given that it is producing some content on its own and creating shows through its network of writers. But to scale and become more popular, it has to create hit shows.

“Blockbusters is what drives any content platform. While we have a good top of the funnel with user-generated content, it is still a hard problem to spot blockbusters.”

Pocket FM has seen encouraging results with its AI deployment. It has more than 40,000 series on the platform with help from AI in terms of voice creation. Plus, the company has generated $3 million in revenue from them. Overall, the platform clocked in $127 million in financial year 2024.

The most difficult challenge for the company is to strike the right balance of AI helping creatives and churning out content quickly. There is always a risk of people using AI to speed up content production and compromise quality. As a result, the platform becomes filled with mediocre content, and it becomes tricky for algorithms to highlight good shows.

Puneet Sharma, an India-based writer and lyricist, pointed out that in a world where so much work is formulaic, artists will have an onus to prove their authentic work.

Sharma added that AI tools can help writers generate ideas and learn different styles. But this means that in the process of using these tools, the process of learning through failure could be lost.

Nayak said that some of its writers and creatives already use AI tools. The idea for the company is to provide tools with the context of stories and the platform.

Pocket FM has raised $197 million in funding across multiple rounds with backers, including Lightspeed Ventures, Tencent, and Times Internet. The company competes on multiple fronts with other content players, such as Audible, Omidyar Network-backed Pratilipi and Google-backed Kuku FM.