aiqadata labelingaudio mlmusic notationmusic transcriptionomraudio to score transcription
Вакансия из Hirify RU Global, списка компаний с восточно-европейскими корнями Для мэтча и отклика нужен Plus
Пожаловаться
78
Хорошая вакансия
развернуть
Роль четко сосредоточена на оценке музыкальных AI-моделей и качестве данных, но условия контракта и объем занятости не указаны, поэтому формат сотрудничества остается неопределенным.
Неясная занятость
Понятные задачи
AI для музыки
Мэтч & Сопровод
Для мэтча с этой вакансией нужен Plus
Подключить Plus
Описание вакансии
Текст:
Hirify AI/Оригинал
TL;DR
Full Stack Producer (Music Notation, Audio ML): Evaluating and improving AI models for music notation and audio through structured QA, expert scoring, and golden dataset development with an accent on OMR, Audio2Score, Autograde, and audio transcription. Focus on comparing model outputs against musical ground truth, labeling specific performance errors, and building benchmark datasets for engineering and product teams.
Location: Remote; Country: Cyprus
Company
hirify.global, part of Muse Group, develops music-focused products involving music notation, audio processing, and AI models.
What you will do
Expert-score every OMR and Audio2Score release during biweekly QA cycles and perform deeper quarterly QA with A/B tests.
Run competitor benchmarking on golden datasets and provide fast sanity checks on model candidates between releases.
Spot-check model outputs against MSCZ ground truth and identify dataset or benchmark inconsistencies.
Build and label the Autograde golden test set by recording mistake types and annotating performances against the corresponding scores.
Compare Autograde with MatchMySound and extend the golden dataset through audio transcription where possible.
Apply the labeling and QA process to adjacent projects such as AutoSync and Audio2Chords as they scale.
Requirements
Strong musical literacy, including the ability to read and evaluate music notation, scores, and audio performances.
Experience judging OMR and audio-to-score transcription output for accuracy.
Familiarity with structured, rubric-based QA scoring methodology.
Ability to evaluate multilingual lyrics, orchestral, ensemble, and solo scores, tabs, and chord notation.
High attention to detail when checking datasets against ground truth.
Ability to provide structured written feedback usable by engineering and product teams.
Culture & Benefits
Remote contract engagement.
Working hours are discussed and agreed upon.
Biweekly, quarterly, and release-based QA cycles are required as applicable.
The Autograde golden test set is needed before the October 31, 2026 MatchMySound cutover.
Будьте осторожны
Если работодатель просит войти в их систему, используя iCloud/Google, прислать код/пароль, запустить код/ПО, не делайте этого. Это мошенники. Обязательно жмите «Пожаловаться» или пишите в поддержку. Подробнее в гайде →