Shantanu Sharma

Working on a project to increase bank’s customer engagement and to reduce attrition. To do so, created a streamlined process to generate monthly recommendations for each user using machine learning. Tested on big data and achieved 80% model accuracy and an overall customer transaction lift of 5%. Reduced the model runtime by 2 hours and delivered this product to multiple banks in India. Developed a machine learning model with 83% accuracy and 71% recall with to increase the banking customer’s average balance. Also, using business intelligence tools created multiple EDA and reports on banking data to visualize and optimize associated project results. Develop algorithms using Natural Language Processing to categorize banking transactions from transaction description. Technologies used are: Spark, Python, HDFS, Scala, Linux.

Created a web application for handling logistics service where drivers, vehicles and multiple deliveries were managed. Tested, deployed and successfully delivered the product to client having 10K+ trucks(vehicles). Developed clustering models and wrote optimized algorithms to assign deliveries to the best possible drivers. Improved performance of the product by implementing an algorithm that creates an optimal schedule of a intercity/interstate delivery for the driver and for tracking the vehicle and driver at each significant location. It was approximately twice as fast than manual scheduling. Technologies used: Python, Django, MySQL, JavaScript.

Analyzed and enriched all Indian schools data by leveraging various REST APIs and data mining using Python, and created a tagging system to obtain more precise results for each school entry. And created a platform to list this processed school data, which was then used by 100+ school counselors. Wrote various scripts to analyze and increase Google AdSense and other Ad networks revenue by 75%. Technologies used: Python, JavaScript, PHP, AWS.

Lucknow, India

Lucknow, India