Hi 👋, I'm Sandeep
I'm a freshmen (class of 2028) at UC Berkeley studying EECS.
I love math, computer programming, robotics, AI and MLÂ
Extracurriculars: Swimming, Running, Biking and playing Piano
I teach children to build robots and compete in FLLÂ
recent reads: Shoe Dog by Phil Knight, Outliers by Malcolm Gladwell, Open by Agassi, Born a Crime by Trevor Noah
 AWARDS and PUBLICATIONS
MATH AMC12 Top 5%, AIME Qualifier(2022-23)
PROGRAMMING USA Computing Olympiad - Gold Division (2022-present)
ACADEMICS AP Scholar with Distinction, National Merit Finalist, High School Honor Roll (2020-2022)
PUBLICATIONS JSR - Journal of Student Research - research paper:
 Smart Path Generation using Model Predictive Control - Oct 2023
SWIMMING WCAL Swimming Finalist (2022)
TRIATHLON 3rd Place, Bay Area Kids Triathlon, Menlo Park 2019
EXPERIENCE
As director and coach of robotics, I designed robotics curriculums to teach programming & design concepts to students from 1st to 8th grade. Organized 8 -10 teams. Assisted in solving FLL Challenge missions. Worked with 30 kids daily over the summer. Had multiple teams win awards at California FLL state tournament.
Developed machine learning models to predict National Teleradiology Program(NTP) demand from east coast hospitals in the VISN network.
Formulated a new time-series evaluation formula to use for demand-predicting models.
Built a server framework to train various time-series machine learning models on real-time data from hospitals, as well as analyze predictions and provide feedback for an increase or decrease in hospital supply.
eVision: Prediction of future Influenza spikes by applying the Long Short Term Memory model on data from Google search trends for flu symptoms and past Influenza data.
Improvised the prediction program by analyzing and fixing abnormalities in the data obtained from Google search trends. This included the translation of foreign language search terms using Google’s translate API.
PROJECTS
Checkout my github here: https://github.com/sunnythecode
https://www.jsr.org/hs/index.php/path/article/view/4952/2351
Independent research paper on Smart Path Generation using Model Predictive Control at JSR(Journal of Student Research). Wrote about using MPC(Model Predictive Control) to generate optimal vehicle trajectories that avoid obstacles. Ran experiments with my own physics simulator in Python, compiled data on various model parameters in order to optimize it. Paper was accepted for publication on Oct 2023.
https://fermusicrecommender.herokuapp.com/
Built an application that recognizes facial emotions from the a picture provided by the user and recommends a song from a user’s given playlist that conveyed the emotion extracted from the user’s image
Clustered Spotify songs into emotional groups(Sad, Happy, Calm, etc.) based on numerical features such as Danceability, Valence, Energy, Tempo, and more.
Trained a Random Forest classifier to predict the group an unseen song would fit into.
The classifier would then sort the songs given by the user(via Spotify playlist) into emotions, and based on the user’s visual emotion, the system would recommend the user’s songs that match the emotion detected.
Blogs
The practical application of this project is to assist stock traders in their buying decisions by using historical data of stock pricing to identify trends and predict future prices. This product was built to provide assurance when making financial decisions.
This project utilizes common machine learning python libraries, as well as public data to produce a trained model able to accurately predict stock prices, in turn, providing assistance when trading and understanding the market. This project specifically used data containing Apple’s stock prices.
Please read the blog post to learn more.
This project has 3 main parts: facial expression detection, music mood classification, and putting it all together for a seamless user experience.
Facial Expression Detection: the model should first identify a face in the image, then learn to classify the expression into sad, happy, calm, or energetic.Music Mood Classification: the model should use song characteristics such as tone or valence to determine what emotion a song represents.
UI: the user interface is made with flask/html and allows the user to input their image and a playlist if they choose. The app should then present the user with the detected emotion and a randomly selected song that corresponds to that emotion.
Please read the blog post to learn more.