- Conducted a deep dive exploratory data analysis: order history, product specific performance, customer and seasonal trends, and consumer-platform interaction/engagement utilizing data collected by Shopify and Google Analytics. - Mapped product sales and customer interaction to optimize consumer engagement and product development. - Defined KPIs involving social media growth, conversion, and revenue goals

- Defined, executed, and presented self-directed projects.- Developed skill in data acquisition, database management, and exploratory data analysis.- Worked with large datasets (e.g. MTA ridership data, CitiBike data) using a variety of tools including regression techniques, machine learning algorithms, and visualization technologies. Tremendous patterns in speech:- Conducted an analysis of the Trump Administration via NLP.- Scraped all articles and transcripts from the Trump White House Press Correspondent archive. - Utilized sentiment analysis, topic modeling, and exploratory data analysis to tell the story of the Trump Administration with collected data.Tools used: Python, spacy, NLTK, pandas, Gensim, BeautifulSoup, seaborn, Plotly, pyLDAvis, Tableau.New York City and CitiBike- Collaborated in a group effort to explore transportation data for insights and trends.- Extracted all CitiBike data and MTA ridership data, from 2014 and on, to implement a Tableau visualization.- Categorized top subway and bike stations and observed change in ridership over time.Tools used: Python, pandas, Tableau, Spark, time-series analysisEveryone's a Critic: Predicting Metacritic Scores- Created a web scraping script to extract details (cast, crew, score) from all films that have been reviewed and scored by Metacritic.- Leveraged data for regression to predict film scores based on cast and crew's previous scores.Tools used: Python, SQL, pandas, Postgres, time-series analysis, scikit-learn, BeautifulSoupDid You See In The Times?- Conducted an analysis of the New York Time Opinions section archive since 1981 with Natural Language Processing technologies and unsupervised learning algorithms - Built a topic model using latent Dirichlet allocation and utilized Vader for sentiment analysis to visualize fluctuations in sentiment value for topics over time.Tools used: Python, spacy, NLTK, pandas, Gensim, BeautifulSoup, seaborn, Plotly, pyLDAvis, statsmodel, Tableau.