Jun 2019 – Mar 2020 · 9 mos · Bangalore

• > Built a multi-class multi-label apparel classification model using Fastai and Pytorch, to automate the manual product onboarding process, by tagging over 55k products a month.
• > Created a collaborative-filtering based recommendation system to identify similar retailers and products to deliver a truly personalized listing experience for 1100+ SMEs.
• > Developed a Reverse image search service using object detection in videos to recommend similar products from our catalog.
• > Designed a real-time stream processing pipeline using Apache Beam and Google Cloud, to optimize analysis and warehousing of orders.

Jul 2018 – May 2019 · 10 mos · Bengaluru Area, India

• Embedded a C++ lib in Python using PyBind11, creating a static library and enabling use of a custom HTTP protocol (HTTP-DIT). This allowed testing of internal services to be possible, along with building an e2e automation testing framework

May 2017 – Jun 2017 · 1 mo · Bengaluru Area, India

• The data science internship for AnalyticSimplified is conducted under the guidance of Mr. Soudhakar Elumalai and Mr. Jaikumar, during this period I worked on a project that involved social media mining, generating an engagement index for brands, recommending the probability of new users liking their products, while completing said tasks I learnt how to interact with APIs, include Natural Language Processing to segregate useful results and making conclusions that are mathematically accurate and statistically sound.