Description
Machine Learning Engineer for Audio Processing
We seek a Lead Machine Learning Engineer responsible for developing and optimizing Machine Learning Models along with technical work including audio data processing, pre-processing, and feature extraction.
The candidate will experiment with advanced Deep Learning techniques for Speech Recognition and Audio processing in collaboration with Data Scientists and Data Engineers.
Key Skills
Strong background in various ML algorithms and proficiency in Machine Learning and Deep Learning models, particularly those applicable to audio analysis like CNNs, RNNs, and LSTMs.
Strong foundation in Signal Processing including the understanding of audio signal spectrograms and its characteristics like sampling, frequency, and amplitude.
Ability to pre-process audio data to remove noise, normalize, perform cleaning steps, and extract and select relevant features such as MFCCs, Zero Crossing Rate, and Spectral Centroid, among others.
Familiarity with audio formats and codecs.
Excellent Python skills, given its prevalence in Data Science and ML (best practices in OOP, libraries, data analysis, file handling, debugging, .
Responsive to feedback and adjustments.
Slack / Jira / Google Docs / Git.
Excellent problem-solving and communication skills.
Leadership skills to guide the team through the project’s lifecycle to ensure the delivery.
Experience
8+ years of total IT experience.
3+ years of experience developing ML / DL models
Proven experience with Deep Learning frameworks such as TensorFlow or PyTorch
Experience using Pandas, Numpy, Matplotlib, Scikit-learn
Expertise using Audio Processing tools and libraries including LibROSA, PyAudio, Wav2Vec, AudioKit, etc. for audio analysis, manipulation, and feature extraction.
Experience working in Agile environments.
Background
BSCS, BSEE, or equivalent
AI / ML Specialization
Nice to have
Experience with MLOps tools and practices.
NLP techniques related to Speech-to-Text, processing, and understanding speech.
Understanding of Pretrained Models on Large Audio Datasets.
Familiarity with the HuggingFace Hubs
Design of experiments or track record of research activities.
Experience in Educational Technology.