The traditional sub-meter reading process is manual and time-consuming, involving trained personnel to survey every house and property to note down consumption metrics. This process has limitations, including high costs and low convenience for consumers. Newer technologies in Computer Vision and Optical Character Recognition (OCR) can be leveraged to automate the process, making it cost-effective and streamlined. A mobile application can be developed to allow consumers to manage sub-meter readings, with images of meters being taken and processed using OCR techniques. The benefits of this approach include reduced costs, increased convenience for consumers, and improved data collection standards. However, challenges such as image quality issues and limited availability of training data need to be addressed. Deep learning approaches like Region Based Detectors, Single Shot Detectors, and Convolutional Recurrent Neural Networks (CRNN) can be used to detect areas of interest and extract digits from images. The Nanonets OCR API allows users to build OCR models with ease, providing a step-by-step guide for training models using their API.