【BLACK MUSIC MONTH】
NPR Music
Tiny Desk Concert
🤎
GENA
 ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄
VoのLiv.eと
DrのKarrim Rigginsのユニット
♬*°
Douwannabwitasstar!?
Theybetterbegladhavetherapy
readmade
Dream a Twinkle
HOWWEFLOW
Circlez
Ono iya ati baba
youtu.be/XYhtkxR5sV8?si=tgP0…@YouTube
Home is essentially dream of everyone. But not of everyone is fulfilled.Dera Sacha volunteers come forward selflessly and gift readmade homes to poor families walking on the teachings and initiative of Saint Gurmeet Ram Rahim ji. #HomelyShelteryoutu.be/fu_4R4Nf__U?si=ROQM…
introducing `usevoiceai`, an open source Typescript package to integrate end-to-end voice interaction into your app. turn your text based apps into voice native apps.
why did i build it?
voice models have made incredible leaps in recent years but the infra to integrate in real apps is still lacking. model providers provide APIs and SDKs and they are great but it’s fragmented. providers have different contracts and quirks. often, choosing one means leaving other. or worse painfully glueing different parts together to make it work.
also, voice doesn’t work well on HTTP. it needs WebSockets and WebRTC to have acceptable latency. i spent the last month going through this journey in the aim for building an experience that is smooth but most importantly model agnostic. i should be able to plug and play with various models because each have their strengths and drawbacks.
what technical problems it solves?
models are already great. this package just augments them on the infra layer.
provides a single unified interface to work with transcription providers and speech providers
handles the web socket layer for
handles STT -> you custom processing -> TTS pipeline
handles Automatic Speech Recognition based turn processing
handles audio recording on web browsers
simple intuitive APIs for good DX
other convenience utilities for both client and server
think of it as the AI SDK for the voice stack. it was hugely inspired by the API design and convenience of the former.
the future is going to be voice heavy. think wearables, custom hardware or any device that doesn’t need a keyboard. voice is going to be the main interface and i didn’t want to start from scratch every time i need to integrate it in any app that i’m building.
it’s open source and you can start testing it right away. if you are someone who is thinking of integrating voice interactivity in their stack, i would love to chat and understand your needs.
roadmap (not exhaustive):
- add more documentation
- add support for more providers
- add support for WebRTC
- add more demos and polish existing ones
- readmade UI components
- performance improvements
find it here - github.com/ankitiscracked/us…
And this men who are saying, I can't simp ooh I can't do this for a woman, those are the rapists
They don't want to be responsible of anything, they want readmade woman, by who?
We can't compare modiji with this pappu. RG is family made readmade politician. Modiji is self made politician. Even lawyer ram jethmalani once said that he won't even keep this rahul gandhi as a clerk in his office. 👇
Hiiiiii mistaken minds...plz pay attention. It's nt dat ur post is really appreciable,rather sometyms I find a readmade draft of watever I mean to target someone with,is available in d form of ur post...Lolz! So need nt put urself on cloud nine for no reason at all.. Lolz again.
Dauda is exceptional. He's already a readmade man. He's not lacking anything and he's contented with what he has . Dauda capacity is far above the status of a gov, let him go for President.