How to use Eleven Labs Python SDK with Foundry?

Ask the Community
1

I saw the DevCon ElevenLabs demos and want to use their Conversational AI in my Palantir Launch project.

I tried installing the library using PYPI (since it wasn’t available through Conda), however AIP Assist tells me: “Note that PyPI packages are not directly supported in Foundry models, as dependencies must be solved from Conda. Make sure the libraries you’re adding are available through Conda channels.”

Any help with how to get the library imported?

2

Hi @tanay - Have you managed to check out the ElevenLabs contribution to the AIP Community Registry?

Once installed, you should be able to work backwards to see how best to implement this.

We’re excited to see your Launch project!

3

Thanks! Went through the repo, got it working in the VSCode Code Workspace, but the iframe isn’t rendering on Workshop.

I have a fairly simple pre-existing repo that’s configured with my API keys, etc. but the issue I’m facing is a compile time error where a conda-build fails.

For context, all I’m trying to do is pass in some ontology object properties to my Conversational AI Agent at runtime through dynamic variables. docs: https://elevenlabs.io/docs/conversational-ai/customization/personalization/dynamic-variables

I have the code set up for it based on their APIs, but when I commit the code and it runs ‘functions-publish’, I get:

 ERROR × failed to invoke command 'conda-build'
    ╰─▶ No such file or directory (os error 2)

This leads me to believe it’s either a) an issue with my environment, or b) the ElevenLabs library posted on PYPI not being compatible with conda stuff?