Solution
Voice and channel measurement for callback and wire workflows
Sonotheia does not verify liveness or identity. The review instrument does not enroll a speaker or compare a caller against an identity. Confirm storage and retention terms for the actual deployment. It measures the voice and the channel in a recorded request and reports what it found, so the person running the callback or approving the wire has evidence beside their own judgment.
A nominal measurement does not establish human origin or authorization. Keep independent callback, identity and approval procedures in the review workflow.
What can a measurement establish?
Sonotheia measures physical properties of the voice and the channel in call audio and reports each one as nominal, deviated or unavailable, with the validity of the recording. A reviewer decides what to do with the result. Sonotheia does not decide whether a voice is synthetic.
How does this fit callback review?
Review the recorded request alongside the institution's callback and approval procedures. A human owns step up, escalate or continue. The instrument does not authorize a payment.
What does the reviewer see?
Shema is Sonotheia's operator-facing readout and dashboard for call-veracity judgment. It presents bounded voice and channel measurements and an evidence record for the reviewer. It does not give a fake/real determination.
How should a deployment be evaluated?
Confirm data handling separately for your deployment: audio storage, result retention, access controls, training use and subprocessors before use.
Frequently asked questions
- Does this authenticate a speaker?
- No. It does not match a caller to an enrolled identity or supply an MFA factor. Legal classification and data obligations require review for the deployment.
- Does this work on phone calls?
- Sonotheia's measurements are benchmarked on simulated wideband, G.711 and AMR-NB chains built from public spoofing corpora. They have not been validated on live telephone traffic. A real trunk adds handset, network and transcoding effects the simulation did not include, so it stays pending until measured in place. Zoom, Teams and WebRTC meeting audio are out of scope.
- What does the analysis API return?
- The analysis endpoint returns a measurement frame: the validity of the audio, voice and channel readings, and a recommended human action. It does not return an identity or fraud determination. Evidence exports are separate from this response.
- Can I upload or record audio in the public demo?
- The public demo uses controlled scenarios. Visitor audio capture is subject to a separate publication gate and is not offered as a verified liveness workflow.
- How do you handle model drift?
- Versioned calibration history records the configuration used for a reading. That history does not establish channel validation or detection performance.
- What is Shema?
- Shema is Sonotheia's review instrument, with an operator-facing readout and dashboard. Its primary intended use is review of a recorded request after the call and before the transfer. Controlled demonstrations are available at shema.sonotheia.ai.
Ready to evaluate Sonotheia for your voice channel?
Request a demo