Code Racers
Web-based Coding & Typing Training Game
Overview
Early-stage programming students often fall behind during live coding exercises due to slow typing speed and unfamiliarity with syntax. This project addresses both issues in a single, interactive environment.
Problem
First-year CS and software engineering students struggled to keep pace with in-class coding activities, limiting comprehension and engagement.
Solution
Developed an interactive platform that combines typing practice with real coding exercises, allowing students to build speed while reinforcing language syntax through repetition and competition.
Technologies
Frontend
Backend
Key Features
-
Structured curricula for Python, Java, and JavaScript
-
Real-time multiplayer “code racing” mode
-
Live WPM and accuracy tracking
-
Language-switchable exercises for comparative learning
Technical Deep Dive
Component Architecture
Designed a modular Vue 3 frontend using a clear separation of pages, layouts, and reusable components. Leveraged slots to create flexible layouts, enabling consistent UI reuse while supporting varied content.
Dynamic Content System
Implemented a centralized “course sections” data model as a single source of truth. This powered both navigation and exercise rendering, allowing dynamic switching between languages and topics without duplicating logic.
Real-Time Multiplayer System
Built a race session manager using Socket.IO to handle:
Player connections/disconnections
Session state synchronization
Real-time progress updates across clients
On the frontend, clients subscribe to session events and render live race progress using percentage-based completion tracking.
User Experience Decisions
Left-side navigation for quick curriculum access
Tabbed interface for switching languages within the same exercise
Clear separation between Practice and Race modes to support different learning styles
Apollo Link
Real-Time Health Monitoring & Clinical Alerting Platform
Overview
Consumer wearables continuously generate health data, but most of it is underutilized in clinical decision-making. This project explores how real-time vitals can be aggregated, analyzed, and surfaced to providers as actionable insights.
Problem
Health issues are often identified only after symptoms escalate, partly due to limited continuous monitoring and delayed interpretation of patient data.
Solution
Built a system that ingests real-time biometric data from wearable devices, analyzes trends using AI-assisted evaluation, and surfaces risk signals to healthcare providers for review and follow-up.
Technologies
Frontend
Backend
Services
Key Features
-
AI-assisted health signal analysis based on real-time vitals
-
Provider alerting system for abnormal patterns
-
Patient dashboard with historical and current health metrics
-
Continuous ingestion of wearable device data
Technical Deep Dive
Frontend Architecture
Developed a modular Angular application with reusable components for patient data tables, alert views, and modal workflows. Structured the UI to support quick triage and prioritization of patients.
Data Ingestion Pipeline
Integrated with Terra API to aggregate biometric data (e.g., heart rate, activity, glucose where available) from multiple wearable sources into a unified backend stream.
AI-Assisted Analysis Layer
Processed incoming health data using the OpenAI API to evaluate deviations from patient-specific baselines. Instead of deterministic thresholds, prompts were designed to identify anomalies relative to historical averages and patterns of the given patient.
Alerting System
Implemented a prioritization system where:
Patient “risk scores” are dynamically updated
High-risk signals trigger alerts in the provider dashboard
Critical cases generate real-time notifications via Twilio
Data Persistence Strategy
Stored both latest and historical vital metrics in MongoDB, enabling longitudinal analysis and improving context for future AI evaluations.
Travel Quest
Quest-Based Adventure App
Overview
Trip planning is often time-consuming and unstructured, especially when trying to balance personal interests with discovering new experiences. This project reframes travel planning as an interactive, goal-driven experience.
Problem
Travelers struggle to build engaging itineraries that align with their interests without spending significant time researching activities.
Solution
Developed a mobile app that generates location-specific “quests” based on user interests, turning exploration into a structured and gamified experience.
Technologies
Frontend
Storage
Key Features
-
Interest-based quest generation
-
Location-aware activity suggestions
-
Gamification system with XP and badges
-
Local progress tracking for offline usability
Technical Deep Dive
Frontend Architecture
Built a lightweight React Native application with a modular component structure to support reuse across quest views, progress tracking, and user interactions.
State and Persistence
Used AsyncStorage to persist quest progress, XP, and badge data locally, allowing the app to function without a backend and maintain state across sessions.
Quest System Design
Designed a flexible quest model that maps:
User interests > activity categories
Location > contextual quest generation
Party Snail
Multiplayer Minigame Party Experience
Overview
PartySnail is a competitive multiplayer party game where players control snail companions racing toward a shared goal. Progress is earned through a sequence of fast-paced minigames, introducing tension and strategy despite the intentionally slow core movement mechanic.
Technologies
Tools
Key Features
-
Multiplayer party system with public and private lobbies
-
3D game environment with custom-modeled assets
-
Sequential minigame structure with persistent match progression
Technical Deep Dive
Multiplayer System Design
Implemented a lobby system supporting both private party creation and public matchmaking. Maintained synchronized game state across players, including progression, scoring, and race positioning.
3D Asset Creation
Learned and applied Blender to create custom 3D assets for the game environment and characters. Integrated assets into the Meta Horizon environment to align with the intended visual style.
UI and Interaction Design
Prototyped UI/UX flows in Figma before implementation. Designed interfaces for lobby management and minigames.
