❓ Problem

A sword fight. A crowded tavern. A betrayal whispered across a table. In a conventional audiobook, one narrator has to carry all three. Great narrators make that work, but there’s another experience available: a cast of characters, a score and a world you can hear.

Harry Potter shows what that can sound like. Its full-cast editions brought together more than 200 actors, original music and location-recorded sound. That’s an ambitious production, rather than a template most authors can easily repeat. Source: Pottermore Publishing.

Even conventional production involves substantial labour. ACX says one finished hour typically takes five to seven hours of work. A ten-hour book therefore implies roughly 50-70 working hours before assuming any extra work for a full cast and sound design. Source: ACX.

The opportunity is to make immersive editions practical for more books. The unanswered question: can AI help deliver a production people genuinely prefer, at a price that works for the author?

✅ Solution

Launch Castscape, a platform that turns manuscripts into immersive audiobooks using AI character voices, music and scene-by-scene soundscapes.

• Upload the manuscript and direct the cast. The platform proposes characters, pronunciations, scenes and voices. The author approves a short audition before generating the whole book, using licensed synthetic voices or performers who have agreed to participate.
• Build a world around the words. A storm rolls in, footsteps approach and the tavern fills with conversation. The production editor can adjust cues, regenerate individual lines and keep the dialogue clear above everything else. Preserve the original text unless the rights holder explicitly commissions an adaptation.
• Let listeners choose the experience. Standard narration for concentration; immersive playback for the full production. Separate narration, score and ambience controls stop every listener being stuck with the same dramatic thunderclap.
• Keep the series consistent. Save the approved cast, pronunciation guide and recurring locations. Book two should sound like it belongs in the same universe.
• Make author approval part of the product. Provide chapter previews, correction requests and a final sign-off before release. Start with a hands-on production service, then automate the steps that repeatedly work.

The mobile demo explores the listener experience: discovery, book details, playback and soundscape controls. It currently uses device speech synthesis and synthesised ambience. The manuscript-to-production workflow described here is the proposed business, rather than an implemented feature of the demo.

📊 Key Numbers

Market context

• The Audio Publishers Association reports $2.43 billion in US audiobook publisher receipts in 2025. That measures the wider audiobook category, not Castscape’s addressable software market.
• Fiction represented 71% of sales, with science fiction/fantasy and romance each contributing 13%. These are sensible places to test character-led productions.
• The same research says willingness to try AI narration fell from 70% in 2025 to 61% in 2026. Cheaper generation alone won’t win the listener. Source for these figures: APA.