A consent-first pipeline for training on real telephony speech

Why real telephony audio

Most speech models are trained on clean, broadband recordings: studio microphones, read speech, quiet rooms. Real business calls look nothing like that. They arrive at 8 kHz over the phone network, compressed by the carrier's codec, full of background noise, interruptions and natural, unscripted speech. A model that has only ever heard clean audio degrades sharply the moment it meets a real call.

Our position is that the only data worth training on for this problem is the data the product actually runs on: real, consented conversations carried over real networks. This paper describes the pipeline that turns those calls into training data responsibly — the consent model, the de-identification, the enforcement that makes both non-optional, and the human review that produces our verified set.

Consent comes first, per client

Nothing enters the pipeline by default. A customer's call audio only becomes eligible for training after an explicit, recorded opt-in at the account level. Until that flag is set, their recordings are used solely to deliver the service they pay for and never flow into research.

We treat consent as a property of the data, not a one-time checkbox buried in a contract. It travels with every segment, and it is checked again at the moment of use — not just at the moment of collection. Consent can be withdrawn, and withdrawal propagates: segments tied to a withdrawn account stop being usable for training.

De-identification before anything else

Phone calls contain personal data — names, numbers, addresses, and, in our world especially, one-time passwords and verification codes. Before a transcript is ever stored or shown to a human reviewer, it passes through a redaction stage that strips:

• Contact identifiers — email addresses and phone numbers in any common format.
• Numeric secrets — any short digit sequence that could be an OTP, PIN or verification code.
• Free-form personal data surfaced during transcription.

Redaction is deliberately conservative. We would rather over-redact and lose a little linguistic signal than let a single credential or personal detail persist in a training set. A segment is only considered de-identified once this stage has run and been recorded against it.

Enforcing consent and redaction in the data layer

Policies that live only in application code get bypassed — by a bug, a migration, a well-meaning script. So we moved the rule into the database itself.

Each speech segment carries a derived trainable property. It is not something a person sets; it is computed from the segment's own state, and it is true only when the segment is both consented and de-identified. Any query that assembles a training set reads from this property rather than re-deriving the rule, so there is exactly one definition of "safe to train on," and it lives in one place.

We go one step further for the verified set. Promoting a label into the gold training set is blocked at the database level for any segment that is not trainable: the write is rejected outright. It is not possible, even by mistake, to mark un-consented or un-redacted audio as training-ready. The guarantee does not depend on anyone remembering to check.

Human verification

Machine transcripts are a starting point, not ground truth. Every candidate transcript is reviewed by a person who speaks the language, in a dedicated review workbench. The reviewer listens to the audio, corrects the transcript, confirms the language, and either promotes it to the verified set or rejects it.

The model's original output is preserved alongside the human-corrected version. That pairing is what later lets us measure, honestly, how far the current model is from human quality on our own data — rather than trusting a benchmark drawn from someone else's distribution.

The life of a segment

End to end, a segment moves through clearly separated stages:

1. Captured from a real call on a consented account.

2. Transcribed by an automatic model to produce a draft.

3. Redacted — PII and secrets removed before storage or display.

4. Reviewed by a native speaker, who corrects and confirms it.

5. Promoted to the verified set — only if it is consented and redacted.

6. Versioned into a dataset snapshot that models train against.

Each stage is auditable. We keep a record of what happened to a segment and when, so the provenance of every item in a training set can be traced back to a consented source.

Governance

The pipeline is designed to align with Nigeria's data-protection framework (the NDPR) and the same principles that underlie comparable regimes elsewhere: lawful basis through explicit consent, data minimisation through aggressive redaction, purpose limitation, and the ability to honour withdrawal. Because consent and de-identification are enforced in the data layer rather than promised in a policy document, compliance is a property we can demonstrate, not just assert.

What comes next

This paper describes the collection and verification system. Future work covers how we version datasets for reproducible training, how we measure model quality against the verified set, and how partners can build on the resulting speech data through an API. We will publish those as the work matures — with results when we have them, and not before.