Hello, I'm
I am a Computer Science graduate and former Research Intern with a consummate skill of working in the IT industry. Skilled in Python, SQL, Microsoft Word, Management, Machine Learning,
Web Development, Databases, and Algorithms. Having been a Distinction Class holder all through academic life,
and possessing strong communication professional skills, I am a Master of Science (MS) focused in Information Systems from Syracuse Schhol of Information
Studies (iSchool).
In my free time, I read historical non-fiction, biographies, memoirs and personality development books. I also write blogs on WordPress that you may find here. I also love exercising, travelling, learning English vocabulary
and learning French. If you ever plan to meet me, you might go back becoming proficient at English. So I look forward to meeting you soon! Au revoir! J'espère que vous passerez une bonne journeé!
1. Leveraged Python, AWS Redshift SQL, Google Analytics, and Looker to analyze user behavior and campaign performance across the Washington Post platforms to improve the average conversion rate of the articles to 3%. 2. Improved workflow efficiency of data pipelines by 52% incorporating Airflow DAGs to streamline the process of updating newsletter and campaign lookup tables for dashboard content management. 3. Built data pipelines for a price uplift model and collaborated with data scientists to suggest an optimal price for the subscribers.• Leveraged Python, AWS Redshift SQL, Google Analytics, and Looker to analyze user behavior and campaign performance across the Washington Post platforms to improve the average conversion rate of the articles to 3%. 4. Improved workflow efficiency of data pipelines by 52% incorporating Airflow DAGs to streamline the process of updating newsletter and campaign lookup tables for dashboard content management. • Built data pipelines for a price uplift model and collaborated with data scientists to suggest an optimal price for the subscribers.
1. Ported the STACKS API Collection Tool (Twitter and Meta APIs) from Python 2.x to Python 3.9 by leveraging the current versions of social media APIs, Python and MongoDB on containerized Ubuntu virtual servers, optimized runtime efficiency by 15%. 2. Processed semantic information of tweets such as hashtags, text, and URL from log files and inserted the prominent tweet information, its date-time stamp, rate limit and delete notification into MongoDB. 3. Updated STACKS persistence layer to utilize native python logging with virtual container volumes to expose its files and abstracting containerized implementation.
1. Perform data analysis and develop key performance indicators (KPIs) using Tableau and R to identify business metrics, infer cause-and-effect scenarios, formulate data-driven solutions, and provide recommendation through dashboard presentation to increase student success rates. 2. Support systems for academic programs by analyzing large datasets using Microsoft Excel, perform ETL, data cleaning, data manipulation and data mining operations, test and validate system functions, and troubleshoot system errors.
1. Co-ordinating with the College of Professional Studies to perform a data visualization project on ‘Online Student Data Analysis’ to develop a more sophisticated ability to analyze students using a series of Tableau dashboard views.
2. Exploring qualitative characteristics of students such as sex, race, religion, and ethnic backgrounds, correlating these data items to determine online enrollment since the COVID-19 pandemic.
I devised an algorithm on the extended application of Bins Approach using statistical moments for Malaria Parasite and COVID-19 Image Classification from Normal and Viral Pneumonia Images using Python. Through the feature extraction algorithm, I passed the 96 feature vector components through the feature selection techniques to filter and perform dimensionality reduction for efficient classification. Later, I evaluated the system with a binary and multiclass classification system for COVID-19/Normal/Viral Pneumonia X-ray images using a pseudo-coloring algorithm and Global Histogram Equalization (GHE) function. With this approach, I procured accuracies in the ranges of 95-99% for the Malaria Parasite Dataset and 88.23%, 88.42%, and 88.34% for the Lung Image Dataset. Finally, I analyzed and examined the feature vectors extracted from the algorithm and its contribution to the detection and classification process of Malaria Parasites using matplotlib and seaborn libraries of Python and Excel Visualization Tools. Under the tutelage of my HOD Dr. Kavita Sonawane, I'm working in furtherance of my earlier research project.
1. Founding and Core Committee Member of TEDxSFIT held on 25th August 2019.
2. Entirely responsible for the Curation of Speakers, Overall Communications, Program Flow and Event Management, Host Script and Valedictory, Documentation, Drafting of Sponsorship Proposals, and Permission Letters for TEDxSFIT.
Relevant Courses: Introduction to Data Science, Management Principles for Information Professionals, Information Management and Technology, Database Administration Concepts & Database Management, Information Poilcy, Applied Machine Learning, Data Warehouse.
Awarded with a Degree of Class I Distinction.
My associated activities and responsibilities have been: PR Executive at the Computer Society of India (CSI-SFIT),
Volunteer at the National Service Scheme (NSS-SFIT), Literary Head and Magazine Editor at the Student Development Cell, RADlab, TEDxSFIT.
The objective of the project is to show customers content that they would like best based on their historical activity. I applied the knowledge of Data Wrangling, Data Visualization and Item-based Collaborative Filtering to get the desired output.
View Project
In this project, Natural Language Processing (NLP) strategies will be used to analyze Yelp reviews data. In this project, I learnt the basics of Natural Language Processing and applied the Natural Language Toolkit or Nltk for short to perform tokenization, feature extraction using Count Vectorizer, difference between Likelihood, prior probability and marginal likelihood. I also trained a Naïve Bayes classifier model using scikit learn and develop a function in python in order to apply it to the Pandas Data Frame
View Project
In this case study, the goal is to train a Deep Network in order to classify images of traffic signs. The dataset contained 43 different classes of images. For this, I had to perform Image Exploration, Preparation (shuffling of training and testing images) and Transformation (conversion to Grayscale and Normalized Images), which was then passed through Le-Net Convolutional Networks for image classification.
View Project
In this project, I used NLP to clean data (Punctuation Removal, Stopword Removal) and using Count Vectorizer to perform sentiment analysis whether the email message is spam or not using Naive Bayes Classifier. For extra processing, I also use TF-IDF to reflect how important a word is to the email in a collection or corpus of emails.
View Project
I predicted trends of Chicago Crime Rates, future predictions and overall insights using Facebook Time Prophet. I used Python over Jupyter Notebook.
View Project
In this project, I've applied my knowledge of Python and the foundations of pandas by answering interesting questions using the Stanford Open Policing Project dataset and the Weather dataset. Note that the project only covers Rhode Island as the district of police activity. In trying to answer such questions, I performed cleaning messy data, creating visualizations, combining and reshaping datasets, and manipulating time series data.
View ProjectI published this research paper 2020 for the 5th IEEE International Conference on Computing, Communication and Automation (ICCCA 2020) on 30th October 2020.
View Publication