- Developed a Finance Chatbot enabling users to inquire about finance data such as overall summaries, specific data points, or pattern-based suggestions, utilizing open-source LLM alongside Entity and Intent Recognition models.- Created a Golang REST API wrapper for clients requiring Golang endpoints.- RealTimeGPU: Transformed password breaking with a Real-time Parallel GPU system, reducing processing time significantly.- Insurance & Investment Systems: Developed comprehensive platforms providing clear insights into policy details and stock market data.- Profile Data Extraction: Engineered a robust system to extract critical profile data from intricate bank statements.- PopplerLambda: Pioneered advanced document processing capabilities.- Monitoring & Analytics: Implemented daily metric tracking, offering insights into user behaviors and processing trends.- Parser Enhancements: Improved parsing accuracy by refining categories and introducing transaction confidence mechanism.- Integrated Deep Learning for sentiment analysis.- Leveraged NER and rule-based techniques for efficient data parsing.- Adept at delivering scalable solutions under tight schedules.

A startup aimed at easing the financial lives of people by providing a 360-degree view of their transactions in one app, including Credit Score analysis and assistance in building credit score.- Developed Natural Language Processing engine to extract financial data from different sources.- Applied basic machine learning techniques to improve data accuracy.- Developed algorithms to process various financial PDFs and extract data.- Created a testing framework for the parsing engine.- Built dashboards for various clients to display user data.- Designed custom data structures to enhance parsing speed and accuracy.- Developed an in-house NLP engine to parse financial statements and alerts. Currently working on building a deep learning engine to increase accuracy and efficiency from 90-95%.

- Working on building financial portfolios using historical correlation between stocks. The objective was to construct portfolios consisting of stocks that exhibit high correlation. Experimented with multiple correlation coefficients and achieved success with the Pearson Coefficient.- Created models utilizing 50+ financial indicators, employing Neural Networks for analysis.- Being a small team, I also worked in developing a basic front end using Django.

Built an automation engine for clients to collect and scrape data using Python, Beautiful Soap.

I took up this job with a call center to manage my expenses while studying. At this job, I learnt how to be a team player and how to communicate with customers to solve their problems, helping me excel in communication.