Blog
All
Company updates
Research
Product
Case studies
Introducing the new Quail VAD 2.0: Robust speech detection for real-time voice AI
Quail VAD 2.0 is a standalone, noise-robust voice activity detector built for production voice AI - a drop-in Silero alternative that holds steady where energy-based VADs misfire.
How telli powers enterprise Voice AI with LiveKit and ai-coustics
At enterprise scale, bad audio becomes expensive. See how telli uses ai-coustics to improve voice agent reliability across millions of calls.
Quail Voice Focus 2.1: New flagship speaker isolation, 10x more efficient
Quail Voice Focus 2.1 brings 10x more efficient speaker isolation to voice agents — without the performance trade-off.
Voice Focus 2.0 + Native LiveKit Integration: Production-ready voice agents are here Copy
Quail Voice Focus 2.0 launches with native LiveKit integration, bringing foreground speaker isolation and reliable audio to production voice agents.
Behind Quail Voice Focus 2.0: Technical benchmark and evaluation
A technical deep dive into Quail Voice Focus 2.0. See benchmarks, training methods, and how foreground speaker isolation improves ASR and VAD.
How Synthesia scaled voice cloning quality by improving audio at the source
When Synthesia reduced training data to 30 seconds, recording inconsistencies became critical. See how audio preprocessing improves voice clone stability.
What Word Error Rate tells us about Voice AI quality in production
In Voice AI, sounding clean isn’t enough. We explain what’s Word Error Rate and why transcription accuracy matters more than perceptual audio quality.
Simplifying audio enhancement for voice AI: Human vs. machine models explained
Speech enhancement for human perception is not the same as for machines. Learn the difference between Rook and Quail and how to choose the right SDK model.
Get started with the ai-coustics SDK in four simple steps
Want to test speech enhancement in your pipeline? This guide shows how to set up the ai-coustics SDK and run your first Quail test.
