AI Product Development Case Study | Anime Smart Toy
Client Profile
A major Japanese animation studio managing a globally recognized, family-friendly anime franchise.
The Problem: Static Merchandise Was Losing Ground, and an Unsupervised AI Toy Risked the Brand
Plush toys and plastic figures were losing fans to digital and connected experiences, and the studio had no safe way to build a premium interactive product to replace that revenue. Static merchandise sales were declining as fans shifted toward digital engagement, but the studio lacked the in-house hardware and AI engineering expertise to build a conversational toy from scratch.
The bigger risk wasn’t the missed revenue — it was the brand exposure. A conversational AI product aimed at children that breaks character canon, or generates an inappropriate response, does lasting damage to a family-friendly franchise’s reputation and its licensing relationships.
| Risk | Business Impact |
|---|---|
| Declining static merchandise sales | Lost revenue as fans move to digital/interactive experiences |
| No in-house hardware or AI expertise | Studio couldn’t build a companion device alone |
| Unconstrained conversational AI | Risk of canon-breaking or unsafe responses to children |
| Brand and licensing exposure | One bad interaction damages a family-friendly IP and its partners |
The Solution: AI Product Development From Hardware to Character Voice Cloning
ISZ.AI led the AI product development process end to end, from concept through manufacturing, turning a static character license into an interactive smart toy with companion-device behavior. The build combined custom hardware, character voice cloning, and strict canon safety controls so the studio could ship an interactive product without losing control of the character.
- Hardware Engineering: We designed a custom PCB, embedded microphones, and a Wi-Fi module housed within a premium plush design, managing the full OEM/ODM lifecycle.
- Character Voice Cloning: We worked directly with the character’s original voice actor to train a secure, high-fidelity TTS model that preserved the character’s exact tone.
- Canon Safety & Governance: We built a constrained conversational AI tuned on approved scripts and character rules, with safety filters and review workflows that cut the risk of hallucinated lore or inappropriate responses.
The Results: A New Interactive Revenue Stream Without Compromising Brand Safety
The smart toy gave the studio a premium interactive merchandise category it couldn’t have shipped safely on its own, and the governance layer meant the studio never had to choose between engagement and brand control.
- High Engagement: The smart toy averaged 45 minutes of daily interaction per active user during launch.
- New Revenue Stream: The hardware product opened a new interactive merchandise category beyond static plush and figures.
- Governed Launch: The content governance architecture supported studio review, approved dialogue boundaries, and launch-window safety monitoring.
Learn More
Explore our related capabilities: