Wearable Heart Rate Monitors are still a technology under research. None of the wearables are able to predict accurate Heart Rates in noisy conditions like exercising or workout. Since most of them are based on PPG sensors which are highly susceptible to noise, so we thought why not leave it up to the machines to decide what kind of processing it wants to do. We plan to make the sensor module using a PPG analog front end from TI and a TI MCU. We will first create our dataset from a huge number of experiments and feed it to a ML algorithm to train offline. We will then use the trained parameters to predict the HR online. The most tricky part of this project will be the feature selection and choosing the most optimum algorithm which can give a near real time performance.