Present: I am a software developer and incoming University of Washington student studying Informatics. I build practical software for education, operations, finance, and student opportunity, with an emphasis on making complex workflows easier to use. My main interests are full-stack engineering, applied AI/ML, product design, and business strategy.
My recent work includes a College Essay Prompts API with 1,700-2,000+ prompts and 500 accounts before sunset, an OpenPath education platform supporting 20+ nonprofits, UW FUEL for meal and macro planning, admissions and financial-aid product work at Cledge and Kollegio AI, and a UW SEAL software proposal tied to a $62K student reimbursement workflow. I enjoy working between engineering and strategy: identifying practical problems, building the system, and refining it for real users.
AI/ML: I have built applied machine learning projects including an Alzheimer's image-classification model, a street segmentation model for identifying obstacles in urban scenes, and AI-assisted education/workflow tools. I am interested in AI systems that are grounded in useful product flows: models that help students, teams, and organizations make better decisions while still being evaluated carefully.
Leadership: I founded and lead StuImpact.org, a student opportunity organization connecting students with programs, events, mentorship, volunteering, and education access. StuImpact.org reflects the same product mindset I bring to software: identify a problem, understand the people affected, and build tools that make progress easier.
At UW, I plan to continue developing as a software developer with a business lens, especially in AI/ML applications, education technology, automation, product engineering, and tools that expand student opportunity.
Designed and built an API for structured college essay prompt access, packaging roughly 1,700-2,000+ prompts behind a simple request flow. Measured request turnaround ranged from about 269-1006 ms depending on request size. The product was aimed at students, counselors, educators, college-advising firms, and startups that needed admissions data inside their own tools. It reached 500 user accounts before I sunset the project because the external API costs outgrew the experiment.
Alzheimer's Detection Model and Model Street Segmentation
Built applied image-understanding projects in JavaScript: an Alzheimer's image-classification model using brain imagery and a street-scene segmentation model for identifying obstacles in urban environments. These projects gave me experience with labeling, preprocessing, interface design, and presenting AI outputs in a form users can understand.
Built UW FUEL as a focused product for University of Washington students: a meal and macro planning web app designed around student nutrition decisions, personal data ownership, and optional AI assistance through a user-supplied key. The project emphasizes a narrow student use case, clear privacy choices, and a usable web interface.
At the UW Sensors, Energy, and Automation Laboratory, I contributed to the Blueberries software proposal effort for a $62K Student Technology Fee project. The proposal focused on UW's student reimbursement workflow, where clearer software could reduce confusion, improve trust, and make financial access easier for students navigating university systems.
OpenPath began as a free educational platform for grassroots nonprofits and has evolved into a multi-tenant learning management system. The current migration rebuilds the earlier Flask version into a Next.js, Prisma, Auth.js, and Firebase architecture with tenant workspaces, encrypted credential storage, course tools, files, assessments, messaging, and role-based access. The platform has supported 20+ nonprofits and has given me experience designing for organizations with limited time, budget, and technical capacity.
Built and organized several public systems across my engineering work: DeskPilot, a student command center with a React/Vite frontend and FastAPI backend; MathNotes AI, an OCR note workspace with a Gemini-powered math tutor; InvestHub, a stock-market learning simulator with portfolio tools and optional AI explanations; and EssayDash, an API-key dashboard for admissions-data access. Together, these projects cover full-stack product engineering, applied AI, data-backed workflows, and user-facing interfaces.
I have been building games since 2019, primarily with Unity and C#. Across personal projects and game jams, I have published 7 games and practiced 3D mechanics, player feedback loops, interactive systems, and iteration. Game development shaped how I think about product feel: responsiveness, clarity, and building things people want to keep using.
May 2024 - Aug 2024 | Remote
Selected for an early-stage startup internship cohort of 20 from 5,000+ applicants. Worked on student-facing education technology systems during Kollegio AI's early growth period, contributing to product and workflow improvements for college-support tools across engineering and product tasks.
Jun 2023 - Jan 2024 | Seattle, Washington
Built Flask, Next.js, React, MongoDB, and Azure-backed applications for college admissions and financial aid workflows. Worked across authentication, AI training interfaces, Chrome extensions, Google add-ons, and user-facing product improvements, gaining hands-on experience shipping features inside education technology products.
Sep 2023 - May 2024 | Remote
Led technical direction for a digital marketing organization, managing interns and building web tools with Python, Flask, JavaScript, HTML, and Bootstrap. Developed automation systems for email campaigns, social media scheduling, and AI-supported content generation while learning how to scope engineering work for non-technical users.
JavaScript, machine learning, image labeling | Oct 2023 - Dec 2023
Built a JavaScript-based model for classifying Alzheimer's-related brain-image examples, reaching approximately 90% precision in project testing. The project helped me practice the end-to-end ML workflow: collecting image inputs, labeling, model iteration, and turning raw predictions into something a user could interpret.
JavaScript, image sensing, machine learning | Oct 2023 - Dec 2023
Developed a street environment segmentation model for identifying different obstacles and objects in city scenes, extending my work in computer vision from classification into spatial understanding and scene parsing.
Risk math, portfolio strategy, modeling | 2022 - Present
Created an algorithmic trading model through four years of experimentation with risk math, portfolio strategy, and simulated/managed testing. The work has been most useful as practice in risk, drawdowns, data quality, and how software can support financial decision-making.
Software: Python, JavaScript, TypeScript, React, Next.js, Flask, FastAPI, MongoDB, Prisma, Firebase, Azure, HTML, CSS, Bootstrap
AI and Data: Machine learning, TensorFlow, Gemini/Cohere integrations, OCR, image labeling, segmentation, AI training interfaces
Product: Full-stack engineering, API design, workflow automation, education technology, information security, developer documentation
Leadership: Team leadership, project management, nonprofit operations, student mentorship, technical direction
Languages: English, Tamil