About Me
Hello! I’m Hrithik Pai, a Computer Science and Mathematics major at UC San Diego, also pursuing a minor in Data Science. I have a strong passion for Artificial Intelligence, Machine Learning, and Software Development, and I'm driven by the potential of these technologies to address complex, real-world problems. My goal is to become an engineer who plays a key role in building innovative products, while continuously growing through collaboration and hands-on learning.
Experience
SAIC
Machine Learning Intern
- Implementing a comprehensive end-to-end MLOps system using Python for model evaluation
- Utilizing ChromaDB and TensorFlow to build an NLP Model that assists employees using an internal tool
Triton Software Engineering (TSE) Club @ UCSD
Vice President of Products
- Oversee the development of 6 projects and manage a team of Product Managers
- Implementing a data visualization site for clients using React, JavaScript, and MongoDB
- Led the integration of Google Cloud into the website for file uploads, to increase processing efficiency
Triton Unmanned Ariel Systems (UAS) Club @ UCSD
Computer Vision Developer
- Added Computer Vision functionality to detect mannequins on the ground with high accuracy
- Developed comprehensive testing systems for Saliency models, leading to a 3% increase in accuracy
- Optimized the drone's path for maximum target acquisition while avoiding collisions using Computer Vision
HP Technology Ventures
Data Analysis Intern
- Used services such as Crunchbase and Surfer SEO to analyze data from over 10,000 startup companies
- Wrote SQL and Python scripts to analyze data and identify investment trends, reducing analysis time by 25%
- Pitched companies to higher management at HP Tech Ventures for potential investment opportunities
Hourglass Ventures
Software Engineering Intern
- Developed a dashboard interface allowing 40+ lending partners to effectively monitor investments
- Leveraged RESTful APIs and relational databases, making the product cross-functional and user-friendly
- Crafted comprehensive documentation, including PRDs, Use Cases, and a feature list, ensuring seamless understanding and collaboration throughout the product development process
TeleSense, Inc.
Data Science Intern
- Worked on a DTW Machine Learning Python script to be implemented in the upcoming product launch
- Used SQL to retrieve and test data from remote sensors to purify and validate data
- Enhanced sensor data analysis workflows by refining data processing algorithms in C++
TeleSense, Inc.
Software Engineering Intern
- Involved in developing a multitiered application for customer usage reports and hotspot identification
- Implemented Java code to retrieve data from the backend PostgreSQL relational database to the application server
- Optimized C++ algorithms within the backend server to improve data processing efficiency
Skills
Projects
SDCTA Data Visualization Website
Built a new website for San Diego County Taxpayers Association to allow users to view information about a specific region in San Diego. Utilized Tableau, MongoDB, and RESTful APIs to allow users to log-in and interact with the data. Increased user interactions by integrating Stripe and allowing users to pay for higher tiers of membership.
Diabetes Prediction
Used Machine Learning methods to predict early signs of Diabetes with an accuracy of 93% using several factors. Trained algorithms such as Random Forest, Regression, Support Vector Machine (SVM), and XGBoost for predictive modeling. Used a custom error rate to modify hyper-parameters to avoid false negatives.
CCIDC
Developed a web application for the California Council of Interior Design Certification, a nonprofit focused on setting professional standards for interior designers. Revamped the application process using React, Typescript, and MongoDB, allowing users to log-in and save their work. Used Google Cloud APIs to manage file uploads in a neat and organized manner.
Publications
The Exploration of Habitable Exoplanets using Data Mining Algorithms and Data Manipulation
View Full Paper Here
Abstract:
The NASA Exoplanet Archive is a dataset that is an extraction from the total sets of data from the Keck, Kepler, TESS, and Gaia observations, where observations show that the observed stellar objects have been determined to possess one or more planets. It is continually updated as more and more exoplanets, or planets outside our own solar system, are discovered and documented. Our first objective was to see how many of these entries were duplicates, which would bring the total number of entries we would work with from 29,283 to 4,259. In previous research, this dataset was filtered by determining which of these exoplanets are inside their Circumstellar Habitable-Zone (CHZ), commonly defined as the range of distance from a host star such that a planet may contain liquid water, a key requirement for life as we know it. However, this calculation was done only for exoplanets with M-type host stars. Over the course of our research, we were able to expand this calculation of the CHZ to exoplanets with host stars of all spectral types. We performed more in-depth investigation of planets with G, K, and M types stars by comparing them to planets in the Planetary Habitable Laboratory (PHL) exoplanet dataset to see how many similarities there are. The PHL catalog used its own set of criteria to define those planets in it as habitable. Using this method, we determined that there were 3 exoplanets with M-type host stars, 0 exoplanets with a G-type host star, and 1 exoplanet with a K-type host star.
Application of Data Mining to Search for Potentially Habitable Exoplanets
View Full Paper Here
Abstract:
Many light years away from our own solar system, over four thousand confirmed planets orbit stars in a fashion similar to our own eight planets and the sun. With the discovery of these planets, called “exoplanets,” comes the question of extraterrestrial life, a concept scientists have been exploring for years. The possibility of exoplanetary habitability relies on a number of factors, such as spectral type, density, and eccentricity, but most importantly: whether the exoplanet in question contains water, the fundamental requirement for life, as we know it, to exist. To determine whether an exoplanet provides the ideal conditions for sustaining this vital ingredient for life, we considered the concept of the Goldilocks Zone, or the circumstellar habitable zone (CHZ)—the range of orbits around a star where liquid water is capable of existing. The research we have been conducting this summer utilizes the public dataset provided by NASA and Caltech and data mining methods, including Python and Microsoft Excel, to identify exoplanets with potentially habitable conditions. The discovery of the exoplanet K2-18b’s water vapor-containing atmosphere was a major part of our research, in which we focused on identifying exoplanets with similar attributes to that of K2-18b, in hopes that they too may be able to retain atmospheric water vapor. After a two-month period, we discovered that 59 exoplanets orbit in the CHZ of their host star. As for the K2-18b ruleset, only 1 planet, K2-3d, satisfies the conditions. We believe K2-3d to have a high degree of similarity to K2-18b, but more in-depth analysis will have to be conducted to conclude its potential to support atmospheric water vapor and life as we know it.
Contact Me
Resume: Available on request at below emails
Personal Email: hrithikp910@gmail.com
School Email: hpai@ucsd.edu
Phone Number: (925) 352-7141