Hi, HN, we’re building Mia & Leo: AI personalities backed by purpose-built duplex speech models for revenue calls.Teams building real-world agents told us that more than 40% of callers hang up within the first 30 seconds. On a revenue call, that meant a lost booking, an unpaid balance, a cold lead, or a customer who does not come back.This happens because humans interrupt, correct dates,...
What if talking to AI felt indistinguishable from talking to a person, even after years of daily interactions? Fluid back and forth, empathetic responses and a consistent personality.
That’s the future we’re building. And we’re fully leaning into the bitter lesson that scalable, data-driven methods win.
Read our blog on the first step towards that: https://www.metavoice.io/blog/bitter-lesson-ai-voice-convers...
We’re releasing access to Speech-1, our conversational speech model designed specifically for customer phone calls (8khz telephony) to reduce call drop rates.Try it: https://tts.metavoice.io Samples: https://www.metavoice.ioVoice AI calls today face >30% call drop rates with current synthetic voices. We want to make this tend to 0 by making AI speech far more human-like and engaging.Current...
https://github.com/vatsalaggarwal/whisper-cli
Whisper CLI is a command-line interface for transcribing and translating audio using OpenAI's Whisper API. It also allows you to manage multiple OpenAI API keys as separate environments.It allows you to get transcriptions and translations by simply running "whisper translate" or "whisper transcribe". You can install it by running `pip install whisper-cli`.It also supports multiple OpenAI keys...
https://twitter.com/vatsal_aggarwal/status/1612536547248836608
Tortoise TTS is the best TTS available today. We built an open-source, at cost, pay per second API for it. The quality of intonation it generates is unparalleled, and we hope our at-cost API will make it easier for people to build on top!This allows folks to run via a single API call - it costs $0.03/query. The WAV file is downloadable, we apply no restrictions.We're open-sourcing all our work...
https://twitter.com/vatsal_aggarwal/status/1604995355468173312
We used text to SQL to build a "Webflow for Web3 Dashboards". We discovered loads of problems in the entire blockchain data space outlined below.Demo: Design your dashboard, type your query in English.code: https://github.com/vatsalaggarwal/blockchain-text-to-sql prompt: https://llmhub.com/2/functions/23/shareWe're open-sourcing everything.There is A LOT of blockchain data (~1.5Tb on ETH+BTC)....
Text to Python `re` regex converter using GPT-3 that:- States assumptions / what it has understood. - Example: `Match any text file.` - >`Match a string that can contain anything at the beginning, but must end with one of ".txt", ".csv", ".tsv", ".rtf", ".log”.` - Lets you test the resulting regex on a string easily GitHub: https://github.com/llmhub/text-to-regexOpen-source prompt:...