What the work actually involves

You receive batches of Punjabi audio or video — conversational recordings, read speech, broadcast clips, sometimes noisy field audio — and produce verbatim or lightly cleaned transcripts against a client style guide. Beyond typing what you hear, you tag speaker turns, mark overlaps, crosstalk, inaudible stretches and non-speech events, and flag code-switching into English, Hindi or Urdu according to whatever convention the guide specifies. A share of assignments is review work: auditing another contributor's transcript or a machine-generated draft, correcting it, and recording why. Where a phrase carries dialectal or cultural weight that a literal transcription would flatten, you're expected to leave a short note rather than silently normalise it.

What the screen is looking for

micro1's screening is AI-led and heavily follow-up driven. It probes whether you can actually hear and write Punjabi at native level — Gurmukhi orthography, correct use of bindi, tippi and addak, nasalisation, tone-bearing distinctions — and whether you can identify regional varieties (Majhi, Malwai, Doabi, Puadhi, Pothohari) from how a speaker sounds rather than from a textbook list. It also tests transcription judgment: what you do with a half-word, a filler, an overlapping turn, or a guideline that contradicts how a real speaker actually talks. Prior AI experience is not expected; demonstrated transcription or language-annotation output is.

Logistics

  • Fully remote, contractor engagement, invoiced hourly or per audio minute depending on the batch.
  • Largely asynchronous: work is pulled from a queue against deadlines rather than fixed shifts, though some projects post windows that overlap US business hours for calibration calls.
  • Volume fluctuates with client demand — treat it as variable-hour supplementary work unless a specific project says otherwise.
  • You'll need a quiet workspace, reliable internet, decent headphones, and a Gurmukhi input method you can type in fluently at speed.

Pay observed in the $8–15/hr range; the upper end tends to go to reviewers and to people who handle difficult audio or less-common dialects. Nothing here is guaranteed.