Projects
Public Customer Voice (PCV)
Microsoft GLobal Hackathon 2nd Prize Winner
Python | BERT Embeddings | BERTopic | Databricks | Azure ML | Reddit APIs | Tableau
Public Customer Voice secured 2nd place in a Microsoft executive challenge. I leveraged my TSAR architecture to analyze social media feedback on Microsoft products, creating a Tableau dashboard for informing product decisions based on public perceptions. Leading a cross-organization team, our goal was to uncover product issues, conduct competitive analysis, and assess new features.
TSAR & BEReddiT
UW Capstone Project sponsored by Meta
Python | BERT Embeddings | BERTopic | Reddit APIs | Plotly | Dash App | Heroku
The TSAR System applies NLP and APIs to analyze internet forum discussions, including topic modeling, sentiment analysis, and relevance measurement. This methodology is implemented on Reddit data as BEReddiT, examining 50+ subreddits with 10,000 comments and 1,000 posts each. The aim: detect conflicts for social media moderators and help users evaluate subreddit culture "toxicity" before joining discussions.
Learn More
Statistical Analysis on Startup Funding Data
UW Applied Statistics Project
R
This study reviews several questions of interest for entrepreneurs as well as investors about Start-up organizations. We performed our analysis using the Kaggle dataset of the Crunchbase 2014 snapshot, which includes approximated 50,000 companies. We found that the average amount of money raised, and the average number of funding rounds, both vary by industry.
We concluded that the average amount of seed money invested is increasing by 13.5% annually. We also found, controversially, that companies that did not have seed rounds are 4 times more likely to have a venture round. Learn More
Visualizing The Hunger Project in Africa
UW Data Visualization
Python | Tableau | Tableau Prep Builder
This project is an interactive visualization centered around illustrating the organizational activities of The Hunger Project (THP) in Africa. We aim to represent the work done by THP and its journey over the years through a visual storytelling approach. The targeted audience includes potential donors to THP, members of THP, and anyone interested in learning more about THP activities.
Learn More
LyriQuest
RIT Capstone
Python
Our project enhances sound cataloging by classifying music and providing lyric descriptions, catering to users' emotions and mood-based music selection. It utilizes song fingerprinting, machine learning, and data science to recommend songs, analyze music type, and draw graphs based on song lyrics. Users gain quick insights into song topics and emotions, aiding playlist decisions. Learn More