Location: [Singapore/ Remote / Hybrid]
Department: Data Science / AI / R&D
Reports To: Chief Scientific officer
Employment Type: Full-time
About the Role
We’re hiring a mid-level Data Scientist to help build signal processing algorithms and machine learning models for automated feature detection in digital time-series data. You’ll operate at the intersection of DSP and applied ML, transforming raw sensor streams into reliable, high-accuracy detection capabilities. If you enjoy building detection algorithms from the ground up, optimizing feature extraction pipelines, and applying ML to real-world signal data, this role is for you. Experience with biomedical or physiological signals is a strong plus.
Key Responsibilities
- Design, develop, and validate signal processing workflows and ML models for automated feature detection in time-series data
- Extract, engineer, and optimize features from raw digital signals using DSP techniques (filtering, resampling, segmentation, spectral analysis, artifact/noise mitigation, synchronization)
- Apply machine learning and deep learning approaches (classification, regression, anomaly/change-point detection, sequence models, or 1D CNNs/RNNs) to improve detection accuracy, robustness, and generalization
- Conduct rigorous model evaluation, ablation studies, hyperparameter tuning, and benchmarking against rule-based or classical baselines
- Collaborate with data engineers and software teams to productionize algorithms, ensuring scalability, reproducibility, and seamless integration into Azure-based pipelines
- Maintain comprehensive experiment tracking, version control, and documentation for all signal processing and ML work
- Translate product, research, or clinical requirements into technical solutions and communicate findings clearly to cross-functional stakeholders
- Stay current with academic and industry advancements in time-series ML, DSP, and applied AI
Qualifications
- 2–4 years of experience as a Data Scientist, ML Engineer, or Research Scientist working with time-series or signal data
- Strong foundation in digital signal processing (sampling theory, aliasing, FIR/IIR filtering, FFT/spectral analysis, windowing, resampling, noise/artifact handling)
- Hands-on experience applying machine learning techniques to feature detection in signal or time-series data
- proficiency in Python and scientific computing ecosystems (NumPy, SciPy, pandas, scikit-learn, PyTorch or TensorFlow)
- Experience developing and running ML workflows in Microsoft Azure
- Statistical reasoning and ability to debug complex signal-ML pipelines
Preferred
- Direct experience processing biomedical/physiological signals (ECG, EEG, EMG, PPG, respiratory, IMU, etc.)
- Experience with advanced time-series ML architectures
- Knowledge of MLOps practices (model versioning, CI/CD for ML, drift monitoring, Azure ML pipelines, MLflow/W&B)
- Publications, open-source contributions, or portfolio projects in signal processing or time-series ML
- Academic or industry background in biotech, digital health, medical devices, neuroscience, or applied research
- A passion for sports and exercise like golf, running, yoga etc