University of California, Berkeley, USA
As a first-year student studying computer science and data science at UC Berkeley, I worked at Infosys in 2016 as a software engineering intern on the India business consulting team. Infosys provided me strong mentors, high-impact problems and responsibilities over our project. Our team worked on blending IoT with deep learning to automate store audits for a F100 company.
Our client faced difficulty in maintaining product placement across stores which ended up hurting product visibility, brand recognition, and competitive differentiation. One example would be if your store purchased soda from a specific distributor, they will typically provide a branded refrigeration unit for the contracted products. However, smaller vendors would often time abuse the lack of oversight and place competing products within those units which resulted in a direct, negative impact on the distributor’s strategy and bottom line.
During my time at Infosys, we worked very closely with our client to understand their pain points and bottlenecks. As an intern, I had a lot of autonomy over the vision of the solution and collaborated closely with the data scientists on our team. We proposed a pipeline for a cheap, scalable, low-touch process for store audits using IoT cameras to capture data within the units and performing live audits using computer vision. Leveraging my past experience at Stanford Medicine on brain tumor classification, I worked on developing a convolutional neural network to classify the products placed within the unit. Whenever our algorithm detected an item that was not a part of our distributor’s product lineup, we could notify that specific vendor to fix the issue immediately. From working with cutting edge technologies released in papers to shipping a PoC to a client, my experience at Infosys was invaluable to my career. The opportunity to explore a different culture while working on high impact projects made the InStep internship stand out from all my past internship experiences.