Voice Agents

Parent: Agentic.

Voice Agents turn speech into structured work: STT, an agent loop, TTS, and often telephony/WebRTC. Under Agentic they compose perception with tools. Whisper-class models may appear as dependencies; the product is the realtime pipeline (Pipecat, LiveKit Agents, Vocode).

Technically the hard parts are barge-in, latency budgets, endpointing, and keeping tool calls aligned with spoken confirmations. Failure modes: agents that talk over users, transcripts without case ids, and voice paths that mutate systems without a typed confirmation.

Prefer LiveKit/Pipecat-style stacks when you need realtime media; prefer Vocode-like frameworks when telephony integrations dominate; treat Whisper as STT substrate and Piper as local TTS substrate unless the pack is truly an agent product.

Features such as Perception & Multimodal Sensing and Customer support / Sales / GTM show where voice lands commercially — still under the same gate rules as text agents.

Inline members for audit: Pipecat LiveKit Agents Vocode Whisper Piper. Features: Perception & Multimodal Sensing Customer support Sales / GTM Voice Agent Platform. Related terms referenced: Frameworks, Observability, Structured Output.

Prefer Whisper for STT and Piper for local TTS when composing a round-trip under a pipeline pack; hosted SaaS TTS remains out of cell unless a pack is added.

For this research catalog the acceptance test is boring and strict: after wiring the pack, can a second person replay the same scenario and get the same measurable outcome? If not, the page is still a brochure. Document the scenario, the expected number, and the log line that proves it. That habit is what separates a deepened taxonomy cell from a landscape screenshot.

Also record what you refused: tools not connected, corpora not ingested, metrics not collected. Refusals keep the cell honest when marketing pages claim universality. A taxonomy essay that only lists winners teaches less than one that names the skip conditions next to the install path.

Keep the essay technical enough that an engineer can argue install order — and short enough that a reader finishes it before opening the right-rail member list. Prefer primary docs and repo evidence over landscape blurbs when the right-rail packs disagree.

Operational discipline for this cell: name the install artifact, the health check, and the rollback. Prefer one golden path documented against a member pack in the right rail over five half-wired demos. When packs disagree, believe the primary repository README and a failing test you can re-run — not the landscape one-liner. Measure latency, cost, and error rate on a frozen scenario before declaring the subcategory “covered.” If you cannot point at a log line that proves the scenario, the essay is unfinished even when the word count clears the audit floor.

Cross-link greedily but honestly: every taxonomy wikilink should change a reader’s next click, and every pack slug should resolve to a real pack. Do not invent members to decorate the prose. Prefer fewer true links over many decorative ones. Keep Monday-shaped typing out of the runtime story; observation and ledger remain senior to any agent UI that retypes reality.