Description

You will drive complex data science and machine learning initiatives focused on time series and multi-modal data streams.

Responsibilities

• Process, augment, and engineer features for univariate/multivariate time series data from sensor streams.
• Integrate heterogeneous data types (time series, images, text, audio, structured) into deep learning architectures.
• Develop and implement self-supervised and semi-supervised learning models for temporal prediction and alignment.
• Build and maintain robust data pipelines for ingesting, cleaning, and synchronizing large-scale multi-sensor datasets.

Required Skills

• Expert proficiency in Python (NumPy, SciPy, Pandas) and C++/CUDA for performance optimization.
• Deep expertise in PyTorch, TensorFlow/Keras, and experience with Jax/Flax.
• Proven experience with sequence models (RNNs, GRU/LSTM, Transformers), CNNs, and GNNs.
• Familiarity with large-scale training on multi-GPU/multi-node clusters, including ZeRO optimization.
• Strong background in signal processing, including Fourier/wavelet analysis and noise modeling.
• Experience with probabilistic methods, optimization (convex/non-convex), and statistical foundations.
• Ability to handle diverse sensor modalities (accelerometers, vibration, audio, etc.) and real-world noise.
• Master's degree in Computer Science, Data Science, AI, or a related quantitative field.
• Minimum of 3+ years of relevant experience in data science or AI.

Key Skills

Education

Master's degree