Tag
A team shares how they built a production voice-agent stack using self-hosted Pipecat, Mai Transcribe 2, a Sarvam conversational LLM, and a Malayalam fine-tuned VoxCPM2 TTS, achieving 20 concurrent low-latency voice streams on a single L4 GPU with an automated WhatsApp post-call handoff.
A practitioner reports that two AI voice agents passed demos but failed on real financial services calls — one hallucinated a fee, another skipped a required disclosure when interrupted. The agent that held up used locked, mandatory workflow steps; the post asks how others test for such compliance failures before launch.
The article announces the 2026 ISCA Distinguished Lecture Series, featuring talks on LLM agents, real-world Speech AI, and speaker recognition & anti-spoofing by experts such as Kai Yu, Joseph Keshet, and Kong Aik Lee. The detailed abstract for Kong Aik Lee's talk covers recent advances in speaker verification and speech deepfake detection, including evaluation metrics and ethical considerations.
This paper analyzes spurious onsets in full-duplex speech LLMs like Moshi and PersonaPlex and introduces an inference-time mitigation method based on causal analysis to suppress them without retraining.
Nari Labs introduces Qwen3-TTS and Qwen3-ASR models, providing high accuracy, low latency, and cost-effectiveness in a free public beta, alongside optimized APIs and services for production deployment.
The Awesome-SpeechLM-Survey repository on GitHub systematically organizes the research lineage of speech language models, including classification frameworks, representative models, training datasets, and evaluation benchmarks. It serves as a knowledge map for understanding the field.