Anurag Verma

WORK HISTORY

EDUCATION

ABOUT ANURAG VERMA

I love building intelligence to come up with innovative solutions to business problems. I have driven key architectural decisions to design systems and ML algorithms, delivering scalability and reliability, to solve the business needs. My experience and expertise span the following areas:/:At Zynga, I have leveraged Machine Learning to build constantly evolving rich user contexts, based on the player\'s progression during the game. Further, I have developed a classification algorithm to recommend prompts to nudge users to increase player engagement by 15%(based on Fogg’s Behavior Model).I have always had an interest in the field of healthcare. I took a sabbatical (leveraging all my savings) to work on developing an AI classification algorithm to classify tumours in digitized images. I have shared the basic building blocks on Github. I trained Inception V3 using Tensorflow Slim.(https://github.com/anuragvermaknn/digital-image-analysis-sulli)I also attended a conference on Emergency Medicine at AIIMS, New Delhi, in 2017. At Bobble App, I led the initiative to curate a proprietary labeled data, and developed a ML algorithm with an accuracy of >80% to recommend personalized content based on multilingual user prompts for a user base of my master\'s thesis at IIT Delhi, I worked in the domain of Example Based Machine Translation, where I used different toolkits such as Moses, Giza++, IRSTLM to generate phrase translations from Hindi to English, and later used the probabilistic models to translate the judicial documents from Hindi to English.//:I am a hands on engineer, and have built web applications from scratch. I love to dig into operational bottlenecks and scalability challenges. At Shipsy, I laid the foundations of the data processing and analytics dashboards, which empowered the customers, such as DTDC (Indian FedEx), to get 100X more visibility into the operations, going as granular as possible to compare operational efficiency between any two last mile riders or any two warehouses:I am an early adopter of Apache Spark. At Zynga, I have developed distributed systems, ingesting data at the scale of 4 TB daily, and further developed analytical applications to drive business functions and initiatives, such as experimentation, personalization, engagement, marketing, and instrumentation.