The work
You will be handed queues of short Tamil audio clips and asked a narrow question: does this sound like a native speaker, and where exactly does it fall short? Ratings cover segmental pronunciation, prosody and intonation contours, stress placement, and overall linguistic authenticity — including whether synthesized or model-generated speech betrays itself through unnatural pausing, flattened intonation, or phoneme substitutions. Each rating needs a short written justification in English that a reviewer who does not speak Tamil can follow, which is why the English writing bar is real rather than nominal.
Expect the rubric to be more specific than your instincts, and expect edge cases: regional variation across Tamil Nadu, Sri Lankan Tamil, and diaspora speech; code-mixed English lexical items; speakers who are fluent but audibly second-language. Part of the job is flagging ambiguity to project coordinators instead of guessing, and applying the same standard on clip 300 as on clip 3.
What the screening looks for
micro1's process is AI-led — a conversational interview followed by task-based calibration. It probes for:
- Genuine native-level Tamil, tested by asking you to describe phonological and prosodic features precisely, not just to speak the language
- Ability to name why something sounds non-native, in English, without vague words like "off" or "weird"
- Consistency: whether your ratings hold up when the same or similar clip reappears
- Willingness to follow a rubric you disagree with, and to say so through the right channel
- Honest availability — inflated hours are a common reason contributors are dropped
Background in linguistics, phonetics, voice work, audio engineering, teaching, or broadcast media strengthens an application considerably. No AI experience is required or expected.
Logistics
Contractor engagement, fully remote, asynchronous within task deadlines. Volume comes in project waves rather than a steady weekly load, so treat this as supplemental rather than primary income unless a specific project says otherwise. You will need reliable headphones, a quiet listening environment, and a stable connection. Pay in the $30–65/hr band reflects rates contributors have reported; actual offers vary by project scope, seniority, and whether the work involves annotation alone or rubric development.