What the work involves
You will be on the phone with an AI voice agent, not a person. Each session you're handed a short scenario — a billing question, a booking change, a complaint — and asked to play it out over an outbound or inbound call in Peruvian Spanish. Afterwards you log what happened: where the agent misheard you, where it looped, where it answered something you didn't ask, where the reply was technically correct but sounded wrong for a Peruvian caller. Sessions are short and repetitive by design; the value comes from many calls with consistent, legible notes.
No domain expertise is required and none is being tested. What matters is that your Spanish is genuinely Peruvian — Lima or regional — and that you speak naturally rather than slowing down and over-articulating for the machine. Testers who unconsciously adapt their speech to be easy for the AI produce useless data.
What the platform screens for
- Accent and register authenticity. Expect to be asked to speak, and to be probed on Peruvian usage versus neutral Latin American Spanish.
- Audio conditions. A quiet room, a reliable handset or headset, and a stable phone line matter more here than on text-based work.
- Feedback that is specific. "It was confusing" is worthless; "it read the amount as ciento cincuenta when I said cincuenta y cinco, then never offered to confirm" is the deliverable.
- Immediate, consistent availability. The engagement is short and calendared; someone who can do three hours daily starting now is worth more than someone stronger who starts in two weeks.
Logistics
Remote, phone-based, 1–3 hours per day for 1–2 weeks, starting immediately. Call windows may be tied to the client's business hours rather than yours, so some overlap flexibility helps. Pay is observed at $8/hr and reflects a generalist testing task, not expert annotation — treat it as short-term paid work rather than an ongoing engagement.