Machine Learning Researcher – Audio

This is a fully remote position, open to applicants in Brazil.

📋 Description

• Conduct research on the quality of audio data for machine learning applications.

• Explore the impact of audio quality, signal characteristics, dataset composition, and localized acoustic challenges on the training, evaluation, and deployment of models.

• Create innovative metrics, benchmarks, diagnostics, and evaluation frameworks to assess audio data quality in ways that predict machine learning model performance.

• Analyze and summarize Protege’s audio collection, ensuring the maintenance of clear and current quality scorecards and metrics for essential speech datasets.

• Develop techniques to directly measure true acoustic properties from the waveform, such as effective bandwidth, spectral energy distribution, high-frequency roll-off, noise, clipping, reverberation, distortion, and codec artifacts.

• Establish workflows to evaluate diarized or segmented speech regions, identifying localized degradations that may be overlooked by file-level averages.

• Design and execute targeted evaluations linking audio quality issues with downstream model performance, including ASR performance, speaker embedding stability, learned speech representations, and synthesis quality.

• Convert research insights into reproducible filtering rules, quality gates, and dataset selection strategies that enhance dataset consistency across training sessions.

⛳️ Requirements

• PhD or a Master’s degree with equivalent experience plus a minimum of 4 years in the industry related to machine learning, audio signal processing, speech technology, computer science, statistics, engineering, or a similar quantitative discipline.

• Demonstrated experience in designing and executing data evaluations, audio analyses, benchmarks, ablations, or slice-based analyses.

• In-depth knowledge of speech/audio data and signal properties, including sampling rates, codecs, bandwidth, spectrograms, reverberation, clipping, noise, and perceptual quality.

• Experience in developing or critically assessing metrics, benchmarks, or measurement frameworks for ML systems, data quality, speech technology, or audio signal analysis.

• Ability to relate low-level signal properties to machine learning outcomes, including model accuracy, robustness, representation quality, speaker consistency, or synthesis quality.

• Proficient in transitioning between research exploration and production implementation: capable of formulating hypotheses, conducting experiments, analyzing results, and translating findings into scalable tools or decision-making rules.

• Exceptional written and verbal communication skills; able to produce concise technical documentation and clearly articulate empirical results.

🏝️ Benefits

• High level of ownership and a proactive approach to tasks.

• Opportunities for collaboration with external partners.

• A resourceful and resilient work environment.