• Leveraged my analytical skills to assist the ESG team in becoming more data-oriented, streamlining their reporting abilities and leading to my eventual permanent move to the team as its first and only technical member• Wrote functions in Python using the Pandas library to clean and manipulate data from employee surveys and visualized the findings with Matplotlib, enriching the breadth of knowledge gained in comparison to previous surveys• Mapped Booz Allen commodity codes with Excel and Python to their respective EPA codes, increasing the accuracy of emissions calculations and was given “Gold Award” by team for my “collective ingenuity” in building this comprehensive analysis process

• Employed AWS to scrape video frames and get their transcripts, allowing for ability to uncover insights from audio• Wrote Python code to logically cluster videos into sensible categories which resulted in more meaningful analysis• Generated visualizations with Plotly and constructed insights reports, leading to clients’ successful content creation

Projects: https://github.com/raltermanplaylist.append() — Spotify playlist builder via collaborative filtering recommendation system• Constructed multiple pipelines to pull playlists, artists, and songs from Spotify using Spotipy• Ran Surprise SVD collaborative filtering algorithm, recommending playlists to pull songs from based on inputted artist(s) and the percentage of songs in each playlist by that artist• Designed frontend website using Streamlit to allow users to seamlessly make and access their own unique playlistsGotta Classify 'Em All — Classification project separating legendary from non-legendary Pokémon• Used BeautifulSoup, Requests, and Pandas libraries to scrape and clean data from multiple websites and the Pokémon API• Performed Scikit-learn train-test split on the data and SMOTE to deal with class imbalance; ran multiple models, of which XGBoost and SVM yielded the best results• Analyzed feature importance to identify most influential attributes, and reverse engineered to observe the misclassificationsNBA Win Percentage Predictor — Multiple regression model predicting season win percentage of an NBA team based on its statistics• Utilized Selenium and Pandas libraries to scrape and clean game-average stats over a 20+ year period from the NBA’s official site• Performed logarithmic transformations on variables deemed necessary, due to basketball trends, and normalized all of the data• Created OLS models and narrowed down the variables used for final model, inspecting for multicollinearity along the way• Checked for normality and homoscedasticity of model with Matplotlib library, and ran Scikit-learn train-test split

• Promoted after one month to lead team of interns in acquiring new users via targeted outreach• Responsible for nearly 150 new users on the site — recognized for performance in company newsletter

• Wrote Python script to fix inaccurate file of approximately 90,000 retail locations, improving productivity of brokers• Scraped location data from retailer websites using BeautifulSoup and Selenium, uncovering business opportunities

• Helped one artist reach nearly 20,000 views on his music video; promoted musical artists through social media marketing and emailing blogs/online publications• Conducted research for employees on royalties in the music business

• Ensured field-side media was directed to the right place and managed throughout the game• Assisted in post-game interview process, including taking quotes from players in the locker room and maintaining the press box