Your ears are the ground truth. You'll listen to short field recordings and document exactly what's audible — instruments, tempo, key, background sounds, crowd reactions — through a structured web platform built for this project. Later phases involve grading AI model answers against what you actually heard.
Responsibilities
Listen to 30–90 second field recordings (headphones, your own quiet workspace) and complete structured annotations: which instruments are present, tempo and key, timestamps of background events (traffic, sirens, applause), scene details.
Compare paired recordings of the same song captured at different locations and judge what stayed the same and what changed.
In later phases: review AI model descriptions of clips and mark what's correct, wrong, or invented, with a brief written explanation.
Attend a paid onboarding/calibration session and a short weekly sync.
Required skills
Strong analytical listening: you can pick individual instruments out of a busy, noisy, real-world mix — not just studio recordings.
Can identify tempo (within a few BPM) and musical key (relative pitch with a reference is fine).
Comfortable with detail-oriented, repetitive annotation work in a web platform — dropdowns, timestamps, short structured notes.
Experience and education
Who we're looking for (any of these backgrounds)
Working or gigging musicians, session players, producers, or mixing engineers.
Music educators, ear-training instructors, band/choir directors, conservatory students or graduates.
Thriveth makes AI data-training work easier to find, understand, and navigate. We replace uncertainty with clear opportunities, realistic expectations, and insights from real application journeys.