Multilingual transcription correction / annotation (EN, DE, HI, RO, NO, DA, SV, NL, FI, PL)

We are looking for native or near-native (C2 level) speakers of the following languages to perform audio-aligned transcript correction tasks: English German Hindi Dutch Swedish Norwegian Danish Finnish Polish Romanian Requirements: Native or near-native proficiency (C2) in at least one of the listed languages Good command of English Strong listening and attention-to-detail skills Prior experience in transcription or ASR data annotation is a plus Knowledge of multiple listed languages is highly preferred Task Description: You will work on a web-based platform where you will be assigned audio recordings along with automatically generated transcripts. Your role is to: Validate primarily any highlighted segments (these are AI-flagged and may or may not require correction) Correct any other errors noticed to ensure full alignment with the spoken content Occasionally peer-review corrected transcripts with no highlights against the audio. Note: Not all highlighted segments are incorrect—your judgment is required to identify actual errors. Compensation: Payment will be based on a hybrid approach: Number of corrections made + Duration of audio processed Rates will be aligned with the provided hourly rate (details will be finalized before contract start). Selection Process: Shortlisted candidates will be asked to complete a paid calibration task consisting of a 30-minute audio transcript. Candidates who meet the required accuracy threshold will be invited to join the project. Project Duration and Intensity: Estimated: 1–3 months Possibility of extension depending on performance and project needs We will be looking for 3-4 hrs availability daily during active task periods.