What the work involves
You'll work through queues of Gujarati audio — likely a mix of scripted reads, spontaneous conversation, phone-quality recordings, and speakers from different regions — and produce verbatim or lightly cleaned transcripts according to a style guide. The guide matters more than your instincts: it will tell you how to handle filler words, false starts, overlapping speakers, numerals, English loanwords embedded in Gujarati speech, and inaudible segments. A second part of the job is annotation: flagging Kathiawadi, Surti, Charotari or other regional features, marking code-switching into English or Hindi, and noting cultural references a model would otherwise mishandle.
You will also be asked to proofread your own and sometimes others' work, and to give feedback when the guidelines don't cover a case you keep hitting. That feedback loop is genuinely part of the role — projects adjust their specs based on what transcribers report from the audio.
What the screen looks for
micro1 runs an AI-led interview before any project assignment. For language roles it concentrates on verifiable specifics: where your Gujarati comes from, which regional varieties you can actually distinguish by ear, what transcription work you've done and in what tooling, and how you behave when audio is ambiguous. Expect follow-ups that go a layer deeper than your first answer — naming a dialect is not the same as describing what marks it. No AI background is required; domain knowledge is the credential here.
Logistics
- Fully remote, contractor engagement, no fixed shifts in most cases
- Asynchronous task queues with milestone deadlines rather than hourly supervision
- Observed pay band of $8–15/hr; actual rate varies by project, experience, and quality scores
- Requires a quiet environment, reliable internet, and good headphones — browser-based transcription tools are typical
- Volume is project-dependent and can be uneven; treat it as supplemental rather than guaranteed full-time hours