HRV Monitoring • Personalized Insights • Anomaly Detection
Your AI-powered HRV and health insights companion.
HRV-4 transforms continuous heart-rate-variability and activity data into understandable health indicators, anomaly alerts, and personalized recommendations.
The challenge
Burnout signals and HRV anomalies are difficult to interpret from raw wearable data.
Wearable devices can collect large amounts of heart and activity data, but users rarely understand what the numbers mean. HRV changes are influenced by sleep, activity, stress, recovery, and personal baseline. Without contextual analysis, unusual physiological patterns can be ignored or reduced to generic advice.
HRV-4 addresses this gap by turning continuous HRV and activity streams into personalized, explainable indicators such as biological age, burnout resistance, processing of stress, general health score, performance potential, and sleep quality.
Hidden anomalies
Unusual HRV behavior is hard to notice when it appears only as raw intervals or disconnected graphs.
Missing context
HRV should be interpreted together with activity and sleep periods instead of being treated as a single isolated value.
Generic feedback
Users need personalized explanations that reflect their own data history, not only generic health tips.
App gallery
Key product screens and user-facing outputs.
Each screen focuses on making model outputs understandable instead of exposing users to raw physiological data only.
Sensor connection and real-time monitoring
The sensor page allows users to scan for nearby Polar devices, connect to the selected sensor, and monitor live heart rate data directly from the application. HRV-4 displays the latest heart rate values together with a five-minute live chart, giving users immediate feedback about their stress score.
For this part of the system, we implemented cross-platform sensor communication using native Swift and Kotlin modules, enabling reliable wearable integration on both iOS and Android. The page also includes a five-minute stress score produced by a doctor-approved algorithm, turning short real-time HR measurements into an interpretable health indicator.
ML-powered health scores and daily anomaly map
HRV-4 turns complex HRV measurements into clear, personalized health insights. On this page, users can track key indicators such as processing of stress, burnout resistance, performance potential, biological age, and general health score in an easy-to-understand format. The app also supports daily, weekly, monthly, and yearly views, allowing users to observe how their health indicators change over time.
To build these model-driven insights, we trained our machine learning models using more than 5000 hours of doctor-approved HRV data, allowing HRV-4 to provide meaningful feedback about recovery, resilience, stress response, and overall well-being.
Activity tracking for smarter health interpretation
HRV-4 allows users to log and track their daily activities such as sleep, sports, walking, work, rest, transportation, and other lifestyle events. By combining physiological sensor data with activity context, the system can interpret HRV measurements more accurately instead of evaluating every signal in isolation. These activity records help our models distinguish between different daily states, improving the reliability of personalized health scores and anomaly detection. The app also uses this information to support helpful notifications such as sleep quality feedback and water reminders, making HRV-4 not only a monitoring tool but also a daily health companion.
Dr. Quack: your personal HRV companion
HRV-4 includes Dr. Quack, an AI-powered chatbot designed to help users better understand their health data. Through this assistant, users can ask questions about HRV, sleep, recovery, stress, sensor usage, and the meaning of their model results. After each measurement, the insights section also provides personalized recommendations based on the user’s latest health indicators, helping them turn raw measurements into practical daily guidance. Instead of leaving users alone with numbers and graphs, HRV-4 offers a friendly companion that explains results, guides the user, and supports healthier daily decisions.
Our team
Built by the HRV-4 project team.

Ekin Şahin
Team member

Elvan Buse Anlı
Team member

Mehmet Emre Öğütlü
Team member

Öykü Bicav
Team member

Tarık Ege Bilsel
Team member



Tech stack
Technologies behind HRV-4
HRV-4 combines wearable data collection, backend processing, machine learning inference, and a responsive web interface.
React
TypeScript
Kotlin
Swift
Spring Boot
Docker
Python
ONNX
Qwen
React
TypeScript
Kotlin
Swift
Spring Boot
Docker
Python
ONNX
Qwen

