Connect the way carriers expect. Then get out of the media path.
Calls set up exactly the way the phone network expects, so the vendors you already use just work.
Once connected, audio flows straight to your agent — not through extra hops that add delay.
Interruptions behave identically no matter which carrier or platform placed the call.
Pull a live human onto any call over QUIC — listen in, whisper, or take over — without rerouting the call or adding a hop.
Fan one call out to as many supervisors and dashboards as you need. Everyone watches the same live audio and state at once, with no extra latency on the call itself.
If they speak SIP (RFC 3261), yes. We run with Tata TTL-IMS, Twilio Programmable Voice, Plivo, Bandwidth, Telnyx, and a half-dozen regional carriers in production today. SBC dialect quirks (User-Agent stripping, PANI requirements, AAA dialects) are wrapped in compliance modes — pick `minimal` for most, `verbose` for legacy IMS.
PCMU/PCMA passthrough by default — what the carrier sends is what the agent gets, no transcode hop. Opus 48 kHz on direct-media browser/SDK calls. We can transcode to G.722 wideband if your carrier negotiates it and you want HD audio.
If it's in our dialplan DSL, yes. If it's in Twilio TwiML or a FreeSWITCH XML dialplan, we ship migration shims that map the common shapes (gather/play/dial/transfer/record) onto our 13-value DialplanAction surface. Custom IVR logic goes into a coding-agent recipe.
Recording is opt-in per tenant. Audio is encrypted at rest, streamed to your S3 / GCS / Azure Blob bucket of choice, or kept on the substrate's ClickHouse tier with signed URLs. Caller disclosure macros are first-class in the dialplan.
Yes. `TRANSFER` and `BRIDGE` dialplan actions take a SIP URI or E.164. The AI side gets a clean handoff event; the human side gets the call. Whisper / coaching / barge-in for the human supervisor are wired through the same control plane.
Telequick ships every modality on the same transport.