Experience & Skills

Work Experience

• Instructional Assistant: DSC30 - Data Structures and Algorithms
  HDSI & Jacobs School of Engineering (UCSD), April 2022 - June 2022
  • Collaborated with instructors and 21 other TAs to reinforce student learning of skills and materials through weekly check-ins, informal assessments, and observations
  • Assisted student learning by providing one on one support during weekly office hours
  • Answered 100+ questions through the online class forum(piazza) and one on one support during weekly office hours
  • Developed multiple comprehensive rubrics for quarterly examinations
• Software Engineer Intern
  Qualcomm Institute - UC San Diego Division of Calit2, January 2022 - March 2022
  • Developed an interactive VR visualization in the Unity game engine to replicate the effects of meditation for stress, depression, and anxiety reduction.
  • Implemented a data pipeline through socket communication to transfer respiration belt outputs from python API to C# in Unity, allowing users to seemlessly interact with the virtual world through respiration.
  • Presented results to program advisors, 28 other teams, and countless interested students to demonstrate the immersive capabilities of virtual reality software

Other Projects and Research

• San Diego County Regional Business Growth Forecasting September
  San Diego Association Of Governments, September 2023 - April 2024
  • Conducted collaborative research with a team of 3 and mentors to forecast annual business creation and closing rates at the ZIP Code level.
  • Collected, cleaned, and processed more than 5 socio-demographic and economic datasets from the US Census
  • Built and tested various model architectures from sliding window ARIMAs to LSTM neural nets using sci-kit learn and pytorch.
  • Developed a LASSO regression model to predict immediate next-year birth/death counts and a less accurate LSTM regression model capable of predicting any number of years into the future
• League of Legends Champion Analysis and Recommendation Dashboard
  University of California San Diego, January 2023 - March 2023
  • Developed an interactive dashboard using D3.js, visualizing various facets of League of Legends champions/characters to assist users in discovering new champions to learn/play
  • Introduced a step by step process to narrow down possible champions, based on user preferences and choices, dynamically updating visualizations to only display information on a subset of champions that satisfy those criteria.
• Personalized Steam Video Game Recommender
  University of California San Diego, September 2022 - December 2022
  • Normalized (json->csv), cleaned, and visualized a dataset of over 4 million implicit user interactions.
  • Scraped video game prices and release dates with BeautifulSoup4 to impute certain missing values.
  • Implemented Bayesian Personalized Ranking and Alternating Least Squares matrix factorization algorithms with the Implicit python library to build collaborative filtering recommender systems.
  • Tuned and finalized an Alternating Least Squares model that correctly identifies whether or not a user owns a randomly chosen game with 89.93% accuracy.
  • Produced a final dashboard allowing users to generate 20 personalized game recommendations filtered by genre tags and release dates.
• GPT-3/3.5 Text-Based Fantasy RPG Game
  University of California San Diego, September 2022 - December 2022
  • Wrote a python script interfacing users with OpenAI’s GPT-3 API, to replicate the text based RPG game experience (Dungeons and Dragons).
  • Utilized GPT-3 to automatically generate unique responses to player actions, in lieu of having a human “dungeon master” or pre-preprogrammed responses, allowing for unbounded story potential.
  • Implemented various player actions (do, say, see, …) and menu actions (save/load_game, display_questbook, …)
  • Crafted customized prompts for each action type in order to harness GPT-3’s story telling abilities while limiting it’s response scope to a text-based RPG structure. (ChatGPT is now capable of this without any prompt engineering D:)
• National Power Outages Analysis and Predictor
  University of California San Diego, March 2022 - June 2022
  • Wrangled large datasets(pandas), cleaning and transforming(sklearn) features to discover and visualize(matploblib/seaborn) trends and patterns within national power outage observations
  • Analyzed unusual discoveries and rigorously evaluated speculations through hypothesis testing
  • Utilized scikit-learn and supervised machine learning techniques to implement and optimize a random forest classifier to predict the root cause for major power outages.
• KNN Image Classifier and Image Processor
  University of California San Diego, March 2021
  • Developed a python program to perform image modifications and classifications
  • Implemented basic image manipulation tools such as RGB color scale modifications, image cropping, and image rotation.
  • Implemented a k-nearest neighbors classification algorithm based on RGB euclidean distances to classify images with pre-labeled training data.
• Jeopardy Analysis
  University of California San Diego, December 2020
  • Performed general analyses on Jeopardy game data, investigating trends in contestant behaviors and the distribution of questions and question difficulties.
  • Utilized hypothesis testing to analyze mid-game trends between winning and losing contestants, giving insight into mid-game factors that increase contestants’ likelihoods of winning
  • Utilized bootstrapping and confidence intervals to compute an amount of winnings, with 99% confidence, that game-hosts should prepare to give out while minimizing losses

Skills