Market Surveillance- Calculations on Big Data for various user stories using Pyspark which helped in time optimized results by fine-tuning spark configurations.- Extracted insights of emails by utilizing various Natural Language Processing techniques to preprocess data and then applied pre-trained BERT Sentiment, BERT Summary, and Topic Modelling in a distributed manner to improve time performance for millions of records.- Developed Data Management System which was generic for all… Show more Market Surveillance- Calculations on Big Data for various user stories using Pyspark which helped in time optimized results by fine-tuning spark configurations.- Extracted insights of emails by utilizing various Natural Language Processing techniques to preprocess data and then applied pre-trained BERT Sentiment, BERT Summary, and Topic Modelling in a distributed manner to improve time performance for millions of records.- Developed Data Management System which was generic for all the RDBMS, Object Relational Method was made use of, which was later re-utilized by different offerings as well.- To identify the fraudulent banks and relation between different features in a dataset, Decision Tree and Random Forrest Classifier were implemented which predicted alerts at Precision (81,93) and Recall (96,73).- For simplifying the analysis of data to detect and alert the malicious activity, an advanced Network Graph was developed using the email dataset which helped in Cluster Analysis.Applied K-means Clustering Algorithm and Louvain Algorithm to perform further cluster analysis on the network Graph.- Built various graphs using data visualization libraries for the understanding of data and designed a pipeline for data preprocessing.EQA Sheet Automation (POC)- Proposed and implemented an automation tool that generates a report which identifies all the errors from the activity log file. - The report consists of daily and weekly errors committed by users in the form of bar graphs and line charts for quality inspection. - The report is then mailed automatically to the common Mail group for analysis, therefore, reducing more than 80% of work when done manually. - To store the past records, RDBMS was used where configuration file with DB connections and table schemas were created. It helped in setting up the database at ease during deployment ofapplication. Show less

JAVA Trainee