Initial problem
Traditional stream donations use robotic text to speech voices that feel impersonal and repetitive. Streamers want more engaging, entertaining donation experiences that keep viewers entertained and encourage interaction. Existing solutions lack personality, voice variety, and tight integration with streaming platforms.
Implementation
AI Donations leverages ElevenLabs voice synthesis to transform donation messages into lifelike, expressive speech. Donors can choose from hundreds of AI voices and trigger soundboard effects alongside their message. The platform is built with a React frontend and multiple Go microservices on the backend, using PostgreSQL for persistence and WebSocket for real time delivery to the stream overlay. Supported by an ElevenLabs Grant.
Outcome
What mattered
Hundreds of lifelike AI voices powered by ElevenLabs
Extensive soundboard library for enhanced stream interactions
Real time WebSocket delivery to streaming overlays
Go microservice architecture for scalable processing
Seamless integration with popular streaming platforms
Supported by the ElevenLabs Grants program
Insights
The AI Donations dashboard with voice selection and donation preview
Technical frame
The technical frame sits at the end on purpose: it shows the engineering substance behind the solution.
Next project
Real time infrastructure at a glance.
View project