We make your skill’s voice interfaces sound better. Rather than using each platform’s default TTS renderer to generate a speech response, The Magic Door’s renderer takes input in a proprietary markup format and generates audio optimized for the targeted platform. You call our API to generate an SSML snippet containing an audio tag we generate. The Magic Door’s renderer does the following:
- gives skills a consistent presentation across multiple platforms
- enables quick iteration on audio presentation through a custom markup format
- gives developers access to all of the AWS Polly TTS voices, which can be mixed within a single response
- removes limits on number of sound effects per response imposed by voice platforms
- supports sound effects and recorded audio in many formats
- normalizes audio volume to match differing requirements across voice platforms
We charge by per API request, per minute of rendered audio, and per byte of stored content. If you’re interested in being a beta customer for the renderer, contact us.
- A patent related to this system is pending.