Data Science @ UC San Diego · Machine Learning · Full-Stack · Data Viz
I turn data into things that ship — from 95%-accuracy ML models to live web apps people actually use.
I'm a Data Science student at the University of California, San Diego, concentrating in Machine Learning and Data Analytics & Visualization. My favorite work happens when a model leaves the notebook and becomes something real — a mobile app, a research paper, a live website. I've taken projects from raw data all the way to the App Store and production web.
- 🎓 B.S. Data Science, UC San Diego — GPA 3.7 / 4.0 · Expected May 2028
- 🧠 Focus: Machine Learning, Data Analytics & Visualization
- 📍 La Jolla, CA
- 💼 Open to Summer 2026 internships — and available for freelance work
Need a website, app, or any digital build? I take on freelance projects on Fiverr.
Create a website, app, or any digital service you may need →
| Project | What it is | Tech | Links |
|---|---|---|---|
| Rewisp | An ambient, on-device memory for your Mac — screens become searchable "wisps" (text only, never pixels); ⌘⇧Space to ask anything later. | Swift / SwiftUI · Python · On-device ML | Site · Code |
| GradeHQ | A free, private grade calculator for college students — find the exact score you need on what's left to land the grade you want. | Next.js · TypeScript · Tailwind | Live · Code |
| Cut | AI-powered calorie & nutrition tracker — snap a photo of your meal and get an instant macro breakdown. | Next.js · TypeScript · AI Vision | Live · Code |
| TritonSpend | Mobile-first personal finance app for UCSD students — budgeting, expense tracking, goals, and AI insights. | React Native · TypeScript · Node.js · PostgreSQL | Code |
| Croptimization | ML precision-agriculture platform at 95% accuracy across 100K+ data points; published as a cross-platform app to the App Store. | TensorFlow · React Native · Pandas | Code |
| GestAR | AR-inspired gesture & ASL recognition enabling IoT control with sub-100ms latency. | Unity · C# · Arduino · Raspberry Pi | Code |
| SmartBins AI | Smart waste bins that classify and sort trash automatically with a vision model + Arduino build. | Computer Vision · Arduino · Python | Code |
| CleanPlate | Food-waste tracker for dining halls that measures and visualizes waste at the source. | Python · Data Analysis | Code |
| Reporter | A simple, accessible way for people to report crime in their community. | JavaScript · HTML/CSS | Code |
UCSD CSES — Open Source Developer · Sept 2025 – Present
- Built TritonSpend in React Native & TypeScript; designed a Node.js + PostgreSQL backend with REST APIs, schemas, and triggers for automated transactions.
- Created interactive spending & goal visualizations and collaborated through open-source code reviews.
Croptimization — Lead Researcher · May 2023 – Apr 2024
- Authored an ML research paper on AI-driven precision agriculture, achieving 95% model accuracy.
- Analyzed 100K+ real-time weather and soil data points with TensorFlow, Pandas, NumPy, and Scikit-learn.
Global Education Destinations — HungerHeroes Research Lead · Oct 2022 – Dec 2022
- Co-authored an international research paper on food waste and hunger.
- Built a food-redistribution website connecting donors with recipients.
- Eagle Scout — Boy Scouts of America (Aug 2021 – Present)
- Robotics Club — President: grew membership 246%, managed $10K in funds, raised $12,325, and led the team to a state championship → World Championship.
- Hackathon Club — President: organized MountainHacks (250+ attendees, 16 sponsors) and led STEM outreach raising $3K+.
| Category | Tools |
|---|---|
| Languages | Python, SQL, JavaScript, TypeScript, Java, HTML/CSS, C# |
| Frameworks & ML | TensorFlow, PyTorch, Pandas, Scikit-learn, React, React Native, Node.js, Express |
| Tools | Git, PostgreSQL, Firebase, VS Code, Android Studio, Jupyter, Overleaf |
| Hardware & IoT | Arduino, Raspberry Pi, VEX Robotics, Microelectronics, Bluetooth/IoT |
- 📧 Email: ybhaverisetti@ucsd.edu
- 💼 LinkedIn: linkedin.com/in/yashmitb
- 💻 GitHub: github.com/yashmitb
- 🛒 Fiverr: Hire me
- 🎨 Figma: figma.com/@yashmit
Source for my personal portfolio — a minimal, monochrome "liquid glass" site built with plain HTML, CSS, and JavaScript (no framework, no build step). Features a ⌘K command palette with fuzzy search, a draggable nav with spring physics, pointer-tracked glass glare, 3D card tilt, magnetic buttons, scroll-reveal animations, animated stat counters, and a persistent dark/light theme that respects your system preference.
.
├── index.html # all sections (hero, about, experience, skills, projects, leadership, contact)
├── assets/css/styles.css # monochrome liquid-glass design system
├── assets/js/main.js # theme, draggable nav, glare/tilt, reveals, counters
└── assets/img/ # images & icons
Live at yashmitb.me.
