About the role
What You'll Do:
- Build and iterate on ML models for time series, audio signal, and sensor fusion problems, under the guidance of the ML Lead
- Build and maintain data preprocessing and feature engineering pipelines that prepare sensor data for training and inference at scale
- Package and deploy models into production, including on edge or embedded compute, with attention to latency and resource constraints
- Run structured experiments to evaluate model accuracy, generalization, and failure modes, and document findings clearly
- Support monitoring and retraining pipelines that track model performance and drift after deployment
- Work closely with firmware, hardware, and app teams to turn model outputs into clear, actionable product insight
Must have:
- You have a B. Tech in Computer Science, Electrical Engineering, or related field from a Tier 1 institute (IIT, NIT, BITS Pilani, or equivalent) - mandatory
- You have 3+ years of applied machine learning experience
- You have a solid understanding of time series analysis and at least one of CNNs, RNNs, or signal processing techniques
- You're proficient in Python and at least one ML framework (TensorFlow or PyTorch)
- You've taken at least one model from prototype to a production or near-production deployment
- Bonus: experience with audio signal processing or accelerometer-based activity classification
- Bonus: familiarity with edge deployment of ML models on embedded compute
- Bonus: exposure to MLOps tools and practices
- Bonus: prior experience in IoT, wearables, or health-tech
Requirements
- You have a B. Tech in Computer Science, Electrical Engineering, or related field from a Tier 1 institute (IIT, NIT, BITS Pilani, or equivalent) - mandatory
- You have 3+ years of applied machine learning experience
- You have a solid understanding of time series analysis and at least one of CNNs, RNNs, or signal processing techniques
- You're proficient in Python and at least one ML framework (TensorFlow or PyTorch)
- You've taken at least one model from prototype to a production or near-production deployment