Voice Agent Lesson: It's Not the TTS, It's the VAD Timing and Endpointing
Kindly-Duty272 · reddit · 2026-10-27
A developer's post-mortem on making phone agents sound like they're listening: the robotic feel came from VAD timing, not voice choice. Slowing endpointing and letting short pauses pass made calls calmer and stopped talk-overs. Adding a tiny acknowledgment filler before the real reply made latency feel invisible with zero processing change. Takeaway: voice quality is about turn-taking—treat endpointing as its own milestone, not an afterthought.
More from coding & agent
- Agents built a custom inference engine: AMD MI350X serves GLM 5.3 FP8 37% faster than NVIDIA B200 — yisongyue · 2026-10-28
- Kiro explains how 3 engineers and agents keep 1,000 PRs a week stable with 80+ automated checks — rseroter · 2026-10-28
- Single vs multi-agent: same task, 4 setups, 3x cost gap measured — latent_lurker8587 · 2026-10-28
- Mapping the stack that gives agents inboxes, phone numbers and wallets — Similar_Confusion534 · 2026-10-28
- Solo dev's plan to work around Meta's WhatsApp error 3441038 with a fresh number — Fit-Stretch-182 · 2026-10-28
- One prompt to a 30-second explainer video: 4 rules that make code-drawing agents reliable — _haiderfarooq · 2026-10-28