We'll talk about DL algorithms such as object detection and GANs that form backbone of our scalable retail solution. We'll describe how we ensured scalability of our solutions through a process pipeline that encapsulates data collection strategies at scale to iterate quickly and adjust to new products and field validation tools. Our talk will cover the GPU variants we used to deploy our scalable solutions both on the cloud and on the edge. We'll also outline some of the strategies we employed and compare edge- and cloud-based GPU performance.
New Consumer Packaged Goods(CPG) & products need not just sit in retail shelves without having a link back to their production lines.Our solutions, through a patent pending process, create a feedback link between retail shelves and manufacturing entities, production lines that manufacture CPG products and distribution channels - to enable processes that solve out-of-stock issues in retail shelves and also understand customer behavior at the shelf level.Evolving from a solution that was intended to solve out-of-stock issues in retail shelves in real time, our current generic CPG product training platform is set to create a global database of CPG products with their respective descriptions including ingredients,nutrition etc. In this session, we will walk you through how an ecosystem was built, with deep learning at it's core. You will get insights on how GPU's have helped in speeding up the creation of the ecosystem. The session ends with what the future of retail holds in terms of maximizing the human experience through empathy and responsibility -Nested distribution networks for redistribution of unsold retail shelf food in low income groups -enabled by AI.