When we launched into building Talk to Me, Goose! We focused on the initial use case to build something to help with communications. How do we support someone who loses their ability to speak? But, as development progressed, given the quality of the output and the ease of use, it became clear that there was another more commercial use case staring us in the face – content creation.
Why content creation?
Given neither of us were developers by background, we relied on YouTube videos and training videos to bolster our knowledge. We watched A LOT of videos. As we did, we realized that a lot of the videos were using, or could have benefitted from using, voiceover narration of AI-generated audio. So, while the use case for supporting someone at risk of losing the ability to speak was primarily the driver, the ability to convert text-to-speech using high-quality AI-generated voices was basically the same for content creators.
If we could make available a large library of high-quality synthetic voices for speech generation to content creators, why not? If we could make it easy to use and create a compelling solution, even better! If we could attract content creators to leverage this solution because the voice quality was high enough, the library of voices was large enough, and the multi-lingual nature of the voice synthesis was attractive enough, perhaps this could help to subsidize the use case for those of us in need of a solution to support communication on a day-to-day basis.
A Dual-Use Tool is Born
Given this realization that the use case was right in front of us, we leaned into it. Why not serve both markets and leverage one to help the other, if we could make it work? The cost to maintain the infrastructure of the app is not free, and the objective of building the app was to make it as accessible as possible to people who need it to support their communication needs.
The interaction model for both use cases is the same. Load in text to convert to high-quality speech in a voice of one’s choosing, and get the output. For content creators, enable them to save the output. For those with communications needs, enable them to use the output. Six of one…For content creators, make it affordable and fast to do iterative content development. For those with communication needs, make it affordable and fast to do speech generation. Half dozen of the other…
The Business Model and Our Commitment
Given the dual use case, the business model became clear. The infrastructure is built on the Google cloud. This provides all of the identity management, text-to-speech conversion processing, security, storage, and transaction processing. As a result, there is a non-zero fixed and variable cost to operate Talk to Me, Goose! The synthesis of the high-quality speech that Talk to Me, Goose! generates requires compute resources, as you can imagine. We would love to serve people with communications needs at a cost as low as possible. Our objective is to leverage the mix of use cases to keep the overall costs low. We want to enable access as wide as possible to people who need it to support their communications needs. We intend to invest any excess revenue into this mission and to non-profits serving the ALS community. There will be more to come on this in the future.