Skip to main content
This topic has been closed for replies.
rkolady
rkoladyAnswer
Community Manager
Nov 5, 2025 7:36 PM UTC

Thank you for joining us at the Hackathon, and for reaching out to us!


At the Hackathon, we showed a demo of our new Responses API that will allow customers to build custom integrations and orchestrations by invoking Moveworks Assistant programmatically.


Within this new realm of interoperability, it's possible for customers to build their own voice integrations. As an example, a developer could use a combination of Voice-to-Text and Text-to-Speech around the API calls to Moveworks -- this would transform the Assistant's textual conversations into a spoken voice experience.


However, this is just one potential option for a custom Moveworks experience that the Responses API will unlock. It is not the core focus of our Responses API roadmap, and is not a native Moveworks Voice integration.


The quality of the interactions may vary, as our models are tuned for textual interactions, and spoken word interactions (even with VTT + TTS) end up being quite different -- it's the difference in how users type versus how they talk.

1 reply

rkolady
rkoladyAnswer
Community Manager
Nov 5, 2025 7:36 PM UTC

Thank you for joining us at the Hackathon, and for reaching out to us!


At the Hackathon, we showed a demo of our new Responses API that will allow customers to build custom integrations and orchestrations by invoking Moveworks Assistant programmatically.


Within this new realm of interoperability, it's possible for customers to build their own voice integrations. As an example, a developer could use a combination of Voice-to-Text and Text-to-Speech around the API calls to Moveworks -- this would transform the Assistant's textual conversations into a spoken voice experience.


However, this is just one potential option for a custom Moveworks experience that the Responses API will unlock. It is not the core focus of our Responses API roadmap, and is not a native Moveworks Voice integration.


The quality of the interactions may vary, as our models are tuned for textual interactions, and spoken word interactions (even with VTT + TTS) end up being quite different -- it's the difference in how users type versus how they talk.