Inside a Voice Data Submission: What Happens After You Hit Record

" Is this actually legit? " is a fair question to ask before you record your voice and translate words for a platform you just found. Here's exactly what...

Dialect Library3 min read
Inside a Voice Data Submission: What Happens After You Hit Record

"Is this actually legit?" is a fair question to ask before you record your voice and translate words for a platform you just found. Here's exactly what happens after you hit record, with nothing skipped.

Step 1: You pick a word or phrase

From your dashboard, you choose an English word to translate or a sentence prompt to record in your dialect. This is scoped to the country and dialect you selected during onboarding — you're never asked to record in a dialect you didn't choose.

Step 2: You translate and record

You translate the prompt into your dialect and record yourself saying it, from your phone or laptop. No studio, no special equipment — just a clear recording of how you'd actually say it.

Step 3: AI-assisted tools process the submission

Your recording runs through transcription and quality-analysis tooling. This is the AI-assisted layer your wallet tokens pay for — it's what turns a raw recording into something structured enough to actually be useful as training data, rather than an unlabeled audio file sitting in storage.

Step 4: Other trainers in your dialect cluster cross-check it

This is the part that keeps quality honest: your submission is checked against other trainers in the same dialect cluster before it counts toward rewards. Nobody outside your dialect community is deciding whether your Igbo or Hausa sounds "right" — the people who actually speak it are.

Step 5: Accepted submissions earn tokens

Accurate recordings that pass review earn tokens. Once full quality scoring is live, payouts come from a pool funded by data subscribers, designed to cover what you spent funding the submission and put you in profit on top.

What's live today, and what's still coming

Recording, translation, and dialect-cluster cross-checking are live now. Full quality scoring and the subscriber-funded reward pool are still being built out, which is why personal submission and review counts aren't shown on the dashboard yet. Every recording made today is real contribution toward the dialect coverage the payout system is being built to reward — nothing you submit is thrown away while that infrastructure finishes coming online.