• Engineered a refined matching algorithm for the Marketing team in Snowflake to link customers from different applications, across multiple search axes using Jaro Winkler Similarity criteria, yielding a 30% boost in subscriber base while mitigating identity theft practices• Designed and implemented an end-to-end workflow to target customers with high credit scores, resulting in 15% improved conversion rates for Sales partners. Collaborated cross-functionally across LOBs (finance, marketing, sales), third-party vendors, and bureaus• Conducted simulations using Databricks to optimize strategies and implemented the solution in Java to assign tailored offers, performed User Acceptance Testing while establishing performance evaluation frameworks, resulting in a $5 million annual profit increase• Enhanced customer experience by establishing API connections to external data sources, optimizing offer delivery, and spearheading initiatives to identify KPIs, compile roadmaps, and collaborate with partners for implementation• Revamped customer classification strategy, utilizing a decision tree model in Python’s Orange framework, leveraging interactions between diverse attributes from various credit bureaus, leading to a significant 15% increase in node purity• Parsed and analyzed 3 major bureau XMLs, involving over 3k variables to extract end nodes and critical information, saving 200 hours of work• Produced Tableau dashboards for elevating business insights for strategic decision-making, earning acclaim from the leadership team

• Leveraged Sampling (bootstrapping, SMOTE, Stratified Sampling) and classification methods (Random Forest classifier, logistic regression, SVM) in Python to precisely identify redundant data within the Healthcare Master Data Management system, resulted in 75% reduction in data stewards processing time.• Implemented Auto ML feature of Azure Databricks to determine the most appropriate hyperparameter tuned classification algorithm• Developed and documented data flows and process maps to support the education of 5k users on MDM processes• Implemented complex SQL (window functions) queries in Oracle SQL Server to extract significant information from Master Data Management system and transformed the data into a machine-readable format using Python which tapered data operations by 70%• Generated a parallel process in Alteryx (ETL) to compare the efficiency of the Levenshtein distance method of Fuzzy string matching with the process in Python, resulting in 50% more processing time than the python process

• Automated KPI reports such as total orders, payment rejections using Python and used Power BI (DAX functions) for trend analysis to make business decisions that reduced processing time from 12 hours to 30 minutes, increased the efficiency by 97% and yielded $1M cost savings.• Automated monthly and yearly controls using Python which resulted in 10 times reduced time consumption• Teamed with cross-functional engineers to configure the use of TRAX (cloud-based platform provided by FIS) to provide secure IT services for cash management operations• Received ‘Excellent’ rating for 2 consecutive years which is only awarded to 2.5% of the employees• Volunteered for ‘Bloom Automation Lab’ and conducted workshops for 50+ freshers on various analytical skills such as Excel, SQL, Tableau, POWER BI, Python and RPA tool (Blue Prism)

• Devised a 90% accurate Naïve-Bayes classifier model to perform sentiment analysis of Twitter tweets• Orchestrated Textblob for processing data and Tweepy for official Twitter API