built a playground to mess with ai voice infra this week, goal is to see how natural and quick i can get it.

memebuilding • prev manifold.xyz, twilio.com

It's a basic websocket server I hand rolled at the moment.

Using elevenlabs for speech to text and text to speech and gpt-4o to generate responses.

Gonna start swapping components and vendors to see how fast I can get it.

Rated 💯 by multiple networks. Fluent in /unglish. Unbothered by /whatday of the week it is.

Allegedly a designer.

feels like everyday I discover a new browser api

Check out whisperx

I’ve been running it locally for speech to text and then just using the in browser SpeechSynthesis API

Lots of room to tune but it’s cool to have everything run local

Impact

@abundance just staked $impact on @jtgi's cast.

Support @abundance's nominees by subscribing to auto-fund their curation.

Opt out of /impact nominations in frame

Working on /impact ||


Abundance Economy book: book.abundance.id

"Exploring AI voice infra for natural, quick interactions