Machine Learning Engineer

• Architect and implement self -supervised learning frameworks (masked prediction,contrastive learning, temporal order prediction, etc.).
  
• Design and train transformer encoders for high -dimensional time -series data.
  
• Build data pipelines and augmentation strategies for large, noisy sequential datasets.
  
• Fine -tune shared backbones for multi -task problems (classification + regression).
  
• Optimize and export models for edge / mobile inference (TFLite, Core ML, ONNX,
  quantization).
• Develop and train time -series models (LSTM, GRU, 1D CNN, Transformer) to detect
  user motion and events.  
• Design data preprocessing pipelines including synchronization, normalization,
  segmentation, and feature extraction.
• Collaborate with Researcher to integrate physics -based features into ML model
  training.
• Perform hyperparameter tuning, validation, and cross -device generalization testing
  across smart devices.
• Build tools for dataset management, labeling, and feature visualization.
  
• Maintain documentation, experiment tracking, and reproducibility logs.
  

Requirements

• B.Tech/M.Tech or Ph.D. in Computer Science, AI/ML related field.
  
• 0 -3 years of experience developing ML solutions for time -series or sensor data.
  
• Hands -on experience with self -supervised or contrastive learning (e.g., SimCLR,
  MoCo, MAE, BYOL, BERT -style pretraining).
  
• Strong grounding in time -series / sequential modeling (Transformers, RNNs,
  CNN -1D).
  
• Solid understanding of data preprocessing and normalization for continuous sensor
  signals.
  
• Strong proficiency in Python, PyTorch/TensorFlow, NumPy, Pandas, and Scikit -learn.
  
• Hands -on experience with sequence models (RNN/LSTM/GRU/CNN).
  
• Familiarity with MLOps tools (Weights & Biases, MLflow) and edge deployment
  pipelines.
  
• Strong debugging, version control (Git), and collaborative documentation habits.