Headline

Machine Learning Scientist at Dreamspan| ECG, PPG & Wearable AI | Ex-Temple | Ex-Samsung R&D | Patent Holder in Digital Health | Deep Learning for Biosignals, Arrhythmia & Chronic Disease Prediction

About

I’m a Machine Learning Engineer specializing in HealthTech, Wearable AI, and Biomedical Signal Processing. I’ve designed and deployed deep learning algorithms for ECG, PPG, and video-based health monitoring, including commercial models on the Samsung Galaxy Watch 6. My expertise spans signal processing, deep learning (CNNs, Transformers), generative AI, and cloud-scale ML pipelines. I’m passionate about building AI systems that empower healthcare and preventive diagnostics, leveraging multimodal biosignals to deliver real-time health insights. Keywords: Machine Learning • Deep Learning • Computer Vision • Biomedical Signal Processing • ECG • PPG • Health AI • Wearables • Generative AI • On-device ML

Location

Chandigarh

AI stack

antigravity

claude code

claude

cursor

gemini

github copilot

Work experience

Machine Learning Scientist at Dreamspan, Jan 2026 - Present · 9 mos, New Delhi, Delhi, India, * Machine Learning Scientist at an early-stage health-tech startup focused on AI-driven biomedical signal processing * Work with ECG, PPG, and BCG signals for cardiovascular monitoring and prediction * Detect cardiac abnormalities: arrhythmias, atrial fibrillation, ST deviations, bradycardia/tachycardia * Build deep learning models (CNNs, RNNs, Transformers) for noisy, real-time biosignal environments * Apply advanced feature engineering: HRV, RR intervals, QT variability, wavelet transforms * Develop sleep-stage classification using bed-based BCG signals via PVDF sensors * Model heart rate, respiration, and micro-movements for non-invasive sleep monitoring * Focus on early diagnosis and continuous monitoring via wearable and ambient sensing systems

Machine Learning Engineer at Samsung R&D Institute India, Jul 2022 - Aug 2025 · 3 yrs 2 mos, On-site, • Audio Data Analysis: Audio analysis using LSTM and MFCC captures spatial and temporal features for tasks like speech recognition and sound classification. MFCCs extract spectral features, while LSTMs model sequential dependencies. This combination enhances feature extraction and classification accuracy in audio-based machine learning applications. • Cycling Activity Analysis: Leading a project aimed at recognizing and analyzing cycling activity through advanced ML algorithms applied to sensor data. This initiative helps users track their activity more precisely, delivering valuable performance insights and supporting health-related goals. • ECG Data Analysis for Health Monitoring: Developing ML models to analyze ECG data captured by Samsung watches. This project focuses on extracting ECG features to derive health insights, aiding in the detection of potential cardiovascular conditions and empowering users with accessible health monitoring.

Intern at Samsung R&D Institute India, Jan 2022 - Aug 2022 · 8 mos, Noida, Uttar Pradesh, India

Machine Learning Freelancer at Fiverr, Aug 2020 - Jan 2022 · 1 yr 6 mos, Chandigarh, India, • Intrusion Detection System: Developed an ML-based intrusion detection algorithm to enhance security protocols by identifying and preventing unauthorized access. This solution leverages anomaly detection techniques to safeguard systems in real-time. • Metal Defect Detection: Designed a metal defect detection algorithm using CNN and YOLO, enabling rapid and accurate identification of surface defects in industrial settings. This solution improves quality control and reduces inspection time, providing a reliable, automated alternative for industrial defect detection.

Education

Bachelor of Engineering - BE, Information Technology, Panjab university

Bachelor of Technology, MOBILE COMPUTING, CHANDIGARH UNIVERSITY

Public links

dhruv536.github.io: https://dhruv536.github.io/Dhruv-Portfolio/