Dec 2021 – Jan 2022 · 1 mo · Mumbai, Maharashtra, India

• Designed dashboards to monitor the financial health of banking institutions and compare them to RBI-recommended values
• Wrote Python scripts to implement pipelines to get the required data for the dashboards from the periodic data published by the banking institution

May 2021 – Aug 2021 · 3 mos

• Worked on CNC tool wear classification prediction model under Prof. Alkan Burga.
• Did literature review of more than 50 research papers in the field and developed a working knowledge in the field of intelligent manufacturing.
• Coded and backtested various deep learning models on the CNC milling dataset by University of Michigan.
• Developed a Random Forest based approach for tool wear prediction on the same dataset with an accuracy of 94.4% and wrote a research paper for the same.

May 2021 – Jun 2021 · 1 mo

• 1) Web scraping company data
• Scraped data of 50,000 companies from zaubacorp.
• Data of each company contained the names of the directors and their DIN codes.
• Clustered companies with common directors.
• 2) Predicting NPA of banks during forbearance
• The utilised financial data of companies from Prowess and default data from CIBIL to make a machine learning model that is trained in the period of 2002-2008.
• The model predicts bank defaults in the period of 2009-2015 and MCA data is used to calculate the NPA percentage of individual banks in a particular year.
• A ran a regression between NPA reported and, NPA calculated against the change in stock prices of a bank to find that the more the banks were hiding NPA, the better their stock performed.

Apr 2021 – Jun 2021 · 2 mos · Remote

• Built a Java Spring Boot application for creating a database of influencer profiles on YouTube using Selenium and YouTube Data API
• Created RESTful APIs to access the data from the database and request the service for new data by triggering different kinds of jobs
• Implemented functionality to persist the collected data in a PostgreSQL database using the Data Access Object (DAO) pattern

Dec 2020 – Jan 2021 · 1 mo · Gurugram, Haryana, India

• Analysed stocks using technical and fundamental analysis according to the client requirements and made stock reports for the same.
• Applied various trading strategies in python like Pitroski F score, momentum, pairs trading and backtested them with historical stock market data.