HeyJaredai-coustics
Building the future of audio intelligence. YouTube channel of ai-coustics, voice AI startup based in Berlin.
- Indexed videos, last 90 days
- 10
- Latest publication
- Sep 16, 2026
- Audience
- ~220 subscribers
- Earliest in this view
- Jul 17, 2026
Latest videos
Tyto dimensions: 6 reasons why your voice agent mishears you (opens the original)
Read excerpt
If you're building in Voice AI, understanding audio isn't optional. Most teams learn this the hard way: a call fails, and nobody can say why. Was it background noise? A second voice bleeding in? A dropped packet? A codec squeezing the signal too hard? We built that audio expertise into a lightweight model. Tyto is our audio insight model - it listens to incoming audio and names exactly which of six failure modes is at play. In this video, we walk through what each of these six dimensions actuall
Tyto in-call analysis code walkthrough (opens the original)
Read excerpt
Noise doesn't knock before it wrecks your call → and your agent finds out the same moment your customer does. Tyto can score a live stream while a call is running, so your agent can react to bad audio during the call. In this video, our Dev Rel, Paweł, shows how to wire this into your pipeline and act on it while the call is still going - nudge the caller, check the connection, or dial back barge-in when it gets noisy. Try Tyto for free in our developer platform: https://bit.ly/4gNl3DC ______ 🔗
Nest Talks #6: How to choose the right ai-coustics model (opens the original)
Read excerpt
Bad audio breaks a voice agent before the LLM even sees it. Here’s how to choose the right ai-coustics model for speech enhancement, VAD, and audio insight across your voice AI pipeline. In this episode of Nest Talks, Mila from our marketing team walks through how to actually choose the right ai-coustics model for your pipeline. We cover: — Quail Multi Speaker vs. Quail Voice Focus — multi-speaker rooms vs. isolating one primary speaker — Voice Activity Detection: VAD Multi Speaker vs. VAD Voice
How to prevent voice agent failures? Post-call batch analysis with Tyto 1.1 (opens the original)
Read excerpt
Post-call analysis with our audio insight model Tyto scores every recording in the batch - noise, reverb, interfering speech, packet loss and more. You get a ranked list of exactly which calls broke and what broke them. No manual listening or guessing required. In this video our DevRel, Pawel runs Tyto on 450 real call recordings and digs into the ones it flags. Test for free today: https://bit.ly/4gDwIDC And read a whole guide: https://bit.ly/4izzbBD -------------- 🔗 ai-coustics links Website:
ai-coustics custom LiveKit plugin: Quickstart walkthrough (opens the original)
Read excerpt
Want ai-coustics' newest voice AI models in your LiveKit voice agent without waiting for the official plugin to catch up? In this quick walkthrough, our Dev Rel Engineer, Pawel shows you how to swap in our custom LiveKit plugin, grab your SDK key, and drop in Quail (speech enhancement), VAD, and Tyto (audio insight) - for new or already-live projects. 🔗 Quickstart guide: https://bit.ly/4xSYbbw -------------- 🔗 ai-coustics links Website: https://bit.ly/4xJQuEk LinkedIn: https://www.linkedin.com/c
Publishing over time
Last 90 days. Choose a month to open its work.
Recurring subjects
Named in the text we hold. One piece can cover several.
Audience
~220 subscribers
Measured Sep 19, 2026
Source's subscribers, not the number who saw an individual piece.
About this data
Counts cover the work we have indexed. Tone needs enough text and a confident classification. Excerpts and episode notes are not full articles or transcripts.
Identity or attribution wrong? Suggest a correction.