Experience & Education

Relevant Coursework: Data Structures, Object-Oriented Programming, Computer Organization, Discrete Structures, Systems Programming, Design & Analysis of Algorithms, Software Engineering, Computer Security, Multivariable Calculus

Honors: Dean's List: Fall 2024, Winter 2025, Spring 2025, Summer 2025, Fall 2025, Winter 2025 | President's List: 2024-2025

• Developed 15+ React components for a full-stack kindness-tracking web app using React, Vite, Zustand, Tailwind, and Firebase, enabling challenge creation, posting, and classroom interaction.
• Wrote backend logic with Firebase RBAC and GeoQueries for location-aware tracking, reducing invalid submissions by ~35% while keeping query times under 200 ms.
• Optimized load times by 40% through code splitting and lazy loading, while maintaining a clean record in weekly agile sprint cycles with no failed code reviews.

• Working on a search and rescue platform, building out the topography pipeline that feeds live terrain data into pathfinding and route analysis for missing person cases.
• Develop React/Firebase dashboard components and AI Insights widgets that show commanders real-time outputs such as clue locations, command post data, and live weather metrics.
• Helped design the agentic AI layer, including probabilistic reasoning, path analysis, and knowledge graph generation, offloading the heavier ML work to the cloud to keep the frontend lightweight.

• Reach out to 100+ corporate leaders across 50+ companies, demonstrating professionalism and social skills.
• Organize and execute 25+ annual technical events, including corporate sponsored CTF competitions, security workshops, and industry talks, growing member participation by 40%.
• Compete in 10+ annual CTF competitions, specializing in cryptography and web exploitation competitions.

Next.js, TypeScript, Supabase, OpenAI API, Voyage AI, pgvector, Tiptap, Tauri

• Building an AI-native product management workspace used by 20+ beta users, supporting 70+ product artifacts across PRDs, strategy docs, acceptance criteria, and tradeoff analyses.
• Architected a RAG-powered retrieval layer that turns uploaded product docs into trusted workspace memory, enabling PMs to ask context-aware questions and draft artifacts grounded in their own PRDs, feedback, and strategy documents.
• Developing a decision workflow that converts messy product context into evidence-backed tradeoff packages and comparison matrices, helping PMs evaluate scope, timeline, stakeholder priorities, and launch risk before committing to a plan.

Python, scikit-learn, XGBoost, Pandas, NLTK

• Engineered an ML classification pipeline achieving 95% accuracy on a 15,000-article dataset, using XGBoost ensemble methods to identify misinformation patterns.
• Built a text preprocessing pipeline with NLTK stopword removal, regex cleaning, and TF-IDF vectorization, improving model F1-score by 12% over baseline.
• Compared 5 classification algorithms using cross-validation and documented results and model selection rationale.

React, TypeScript, Vite, Three.js, Node.js, OpenAI API, Supabase, Deepgram

• Collaborated with a team to build an AI-powered mock interview platform with a 5-stage React Router flow, generating personalized live voice interviews from resumes and job descriptions with STAR-based and body language evaluation.
• Built an audio-reactive particle visualization using an 8,000-particle Three.js engine with pulsing dynamics tied to speaking state, delivering a polished visual experience that reacts in real time during interviews.
• Engineered a real-time voice pipeline using voice activity detection, dual streaming STT (Deepgram + Whisper), and GPT-4o for seamless live transcription and low-latency AI responses.