Unstructured Data Scientist - Beyond Limits Talent Pal

• Design, train and deliver machine learning solutions to a wide range of datasets: from small datasets to big data, both unlabeled and labeled.
• Collaborate with the development team and ML engineers to implement, test, and validate new algorithms and integrate the algorithms with the existing front-end and back-end systems.
• Build state-of-the-art models for various modeling tasks (e.g., speech recognition, optical character recognition, text-to-speech, facial recognition, time series forecasting, … etc.)
• Optimize and automate model training and testing, experimentation tracking, development, and production.
• Skilled in breaking down problems, documenting problem statements and estimating efforts.
• Mentor junior engineers, perform code reviews and document design decisions.
• Create web services/APIs for serving ML/AI model results.
• Familiar with MLOps Tools (e.g., MLFlow, Kubeflow, … etc.).
• Familiar with unsupervised, semi-supervised, and self-supervised techniques
  

Minimum Qualifications

• B.Sc. in Computer Science, Electrical Engineering, or similar field
• 2-4 years of experience with one modern language such as C++, Python including object-oriented design.
• 2-4 years of experience with machine learning tools and deep learning algorithms and techniques, including but not limited to:
• PyTorch, TensorFlow, Keras, Kaldi
• CNN, LSTM, RNN
• Scikit-learn, pandas, NumPy
• Optimization techniques and fine-tuning models