KAVITHA VENKATESAN

Passionate Data Science enthusiast adept at leveraging analytical prowess, statistical knowledge, and programming expertise to drive impactful solutions in a dynamic, data-driven environment. Ambitious, career-focused job seeker, anxious to obtain an entry-level position to help launch career while achieving company goals.

• AI For India 2.0 (GUVI)
• Business Intelligence using POWER BI (SKILL Nation)
• Develop NLP solutions with Azure AI Services
• Develop solutions with Azure AI Document Intelligence
• Build an Azure AI Vision
• Train and deploy machine learning model with Azure ML

YouTube Data Harvesting  & Warehousing using SQL, MongoDB and Stream lit 

          Involves  - extracting channel details from YouTube, 

                          -  storing the unstructured  data into  MongoDB Atlas, 

                          -  migrating it into SQL in a structured format as different tables.    

          Organized data is visualized with data visualization via Stream lit application.                       Credit Card Fraud Prediction

           Classification problem with highly imbalanced dataset.

         Predicts fraudulent credit card transactions by Machine Learning    models 

          Accuracy achieved >LR-0.95%,SVC-0.95%,DT-0.92%                           

 Hate Speech Detection

          Aims to analyze and mitigate hate speech in online platforms

          Approaches  include

                            Contextual Analysis, User Profiling, Geospatial Analysis, 

                            False Positive/Negative Analysis 

          Model built using Tensorflow library with keras

          Test loss is 0.077 and Test Accuracy is 97.26%