Ami Mayur Shah
College Park, Maryland – ***** adj5ns@r.postjobfree.com +1-240-***-**** www.linkedin.com/in/ami02/ EDUCATION
University of Maryland, College Park Dec 2020
Master of Science in Information Systems GPA - 3.77/4.0 K. J. Somaiya College of Engineering, Mumbai, India Jun 2019 Bachelor of Technology in Information Technology CGPI – 8.59/10 PROFESSIONAL EXPERIENCE
Synopsys Inc. Jun 2020 – Dec 2020
Data Analyst Intern – Software Integrity Group
• Wrote SQL queries for data extraction and analyzed customer datapoints associated with demand, sales and churn in the supply the chain pipeline - to capture spending patterns using Tableau to improve existing pricing models
• Extracted actionable insights from broad, open-ended questions to influence sales strategy and drive roadmap decisions. Created and managed reports on client data to better manage accounts
• Built costing models to determine true cost, overhead, analyze unallocated cost and profitability
• Used Tableau for data visualization of revenue analysis to model costing and revenue for personalized client targeting University of Maryland, College Park – Graduate Assistant Dec 2019 – May 2020
• Graduate Teaching Assistant for Data Analysis course in Spring 2020
• Graduate Research Assistant for large scale data extraction, sanitation, processing and storage in Fall 2019
• Research Assistantship involved writing SQL and NoSQL queries to pull and aggregate data in various formats. TECHNICAL SKILLS
• Programming languages: Python, R, Scala, SQL, Java, C, C++, HTML, CSS, React (Web Development)
• Database Technologies: MySQL, PostgreSQL, MongoDB, MySQL, Firebase, Oracle, Redis
• Big Data Frameworks & Analytics: Hadoop, Tableau, PowerBI, Apache Spark, Google Analytics, RapidMiner, Adv. Excel
• Machine Learning Frameworks: PyTorch, Tensorflow, Keras, Matplotlib, Numpy, RStudio
• Additional Skills: Amazon Web Services (AWS), Google Cloud Platform (GCP), Salesforce, ETL, Deep Learning PROJECTS
Video Games Sales Analysis May 2020 – Jun 2020
• Determined the frequency distribution of video gaming platforms using Apache Spark’s key-value RDD
• Performed big data complex join operations to analyse top gaming platform performances in various regions
• Performed growth forecast by finding total sales evaluations associated with each gaming genre and their platforms Airbnb Austin Market Analysis Apr 2020 – May 2020
• Created data pipelines using Python and performed data profiling using R programming to predict real estate profitability for investors
• Designed Bagged Trees and KNN models and performed cross validation using Python and Tensorflow
• Implemented Ridge and Lasso regression to assess and mitigate multicollinearity and to perform variable selection
• Produced reports with inferences using standard statistical tests to answer research questions Topic Modelling - Obama Trump Tweets Oct 2019 – Dec 2019
• Extracted a live stream of Obama-Trump tweets using Python and created a data pipeline for sanitation, processing and storage of tweets for topic modelling and exploratory analysis
• Developed deep learning model to analyze topic sentiments to find contrasts between the two tweet categories
• Implemented data visualization using Tableau to create reports about public opinions, behavior and attitude based on individual models
Movie Recommendation Engine Jun 2018 – Jul 2018
• Developed 2 recommendation systems using Python and Keras framework and trained them on a benchmark MovieLens dataset containing 1 million records
• Created and tested a Content-Based Model on movie genres and Collaborative Filtering System on user ratings with high accuracy.
• Improved model performance using sparse matrix factorization and ensemble deep learning models Semantic Modelling for Analysing Sentiment Mar 2018 – Apr 2018
• Designed a data pipeline in Python to perform data sanitation and process tweets to predict sentiment inclination using Random Forests and SVMs
• Built and deployed multilayer perceptron network and 1-D CNN to capture semantic structure
• Improved representation of target groups by more than 20% by utilizing trained word embeddings