Audio input widget

Propose an Improvement
1

Inspired by the “tabs” video and by gpt enterprise voice mode, creating an input widget that allows for audio capture to be stored and/or processed by an LLM could be powerful in building AI-enhanced user experiences.

An example use case could be voice to automatic form filling as you speak. It could also enable “alexa” type voice control of a workshop app like IFTTT.

https://www.linkedin.com/posts/granthoffecker_buildwithaip-palantir-activity-7258986101183328256-oLaz?utm_source=share&utm_medium=member_desktop

2

Hi,

We recently released an Audio Recorder widget that you could use to record audio, listen back to the recording and save it to a media set within your Workshop apps. The audio recordings are saved directly to the media set, so you could run processing (e.g. transcription) or pass them into an LLM in foundry afterwards as usual.

This feature is currently in beta, and we are actively exploring the streaming workflow that you’ve proposed as well as other improvements!