Spatial Computing learning Platform for Medical Students

• Developed an innovative VisionOS application for medical students, featuring 3D organ visualization and interactive quizzes.
• Utilized Swift, SwiftUI, RealityKit, and ARKit, with a MySQL backend and RESTful API for efficient data management.
• Collaborated with a multidisciplinary team to ensure accuracy and educational value.
• Presented the project at the computer science program's 50th anniversary, demonstrating its potential impact on medical education.

iOS Voice-Controlled Charcoal Sales Assistant for Blind Elderly Entrepreneurs

• Currently building a voice-first iOS app with SwiftUI and SFSpeechRecognizer for blind entrepreneurs, integrating Claude API with MCP servers.
• Implemented ConvexDB for real-time inventory sync and voice-controlled cash tracking ("record sale: 2 bags, $40").
• Achieved 100% hands-free operation using AVAudioEngine and AVSpeechSynthesizer for continuous audio feedback.

TestBuilder Platform

• Developed a comprehensive test management platform using Next.js 14, TypeScript, and Supabase, enabling organizations to create, assign, and evaluate assessments through an intuitive interface.
• Implemented real-time test building functionality with drag-and-drop module management, dynamic question types (multiple-choice, true/false, open-ended), and automated grading capabilities.
• Developed a scalable testing platform with group assignments and customizable templates using Tailwind CSS and shadcn/ui, implemented React API for notifications and secure authentication, delivering streamlined experience for administrators and test-takers.

MediAgent - Medical Device Sales Platform

• Developed a comprehensive SaaS platform for medical device sales teams using Next.js, TypeScript, and Supabase, featuring interactive dashboards with real-time sales metrics, lead tracking, and revenue forecasting for products like WoundTrack Pro.
• Implemented an AI-powered lead generation system that identifies and qualifies healthcare prospects based on specific criteria, with detailed lead profiles containing company information, contact details, and match scores to optimize sales targeting.
• Created a personalized email generation system that automatically drafts outreach messages based on lead data, while building a responsive UI with Tailwind CSS and shadcn/ui components for seamless cross-device experience.

AI-Powered Presales Management Platform (Neoris)

• Developed a full-stack AI solution to streamline presales software workflows, leveraging PostgreSQL, Supabase, Langchain, and vector embeddings with pg_vector for RAG (Retrieval-Augmented Generation), integrated with Ollama for LLMs.
• Built a robust frontend using Next.js and TypeScript, while focusing on advanced backend engineering with PostgreSQL, including procedures, triggers, RLS, and Supabase Storage for file management.
• Collaborated within an Agile team to enhance performance and security, actively participating in sprint planning, iterative development, and team reviews. Integrated Microsoft Teams API for seamless collaboration.

Recursive AI

• Engineered the frontend interface of a startup platform for AI Agents to create a hierarchy of tasks, delegating to multiple agents such as Sales, Recruiting and General.
• Implemented a detailed view of each AI Agent's status and data using React, TypeScript, Next.js 15, and React Flow.
• Built an intuitive drag-and-drop interface for task management and agent configuration, with real-time updates and state management.

PlayVision Landing Page

• Developed and Designed PlayVision's landing page using Next.js, TypeScript, and Tailwind CSS, featuring interactive team cards and dynamic scroll-based animations.
• Implemented key components including an auto-playing video showcase, custom UI elements with gradient animations, and a responsive navigation system.
• Optimized performance through Next.js Image optimization and lazy loading while maintaining a modern, responsive design across all devices.

Github Issues Clone

• Created a Fullstack Github Issues Clone using NextJS 14, Supabase with PostgreSQL as a Backend Service utilizing, added realtime commments, AI semantic search for tickets, added file storage for view images with previews, multitenancy for different tenants, implemented Row Level Security to ensure data safety

Regal Rexnord Risk Learning Website

Developed a workplace safety training platform featuring Unity/C# educational games integrated with a React/Material-UI analytics dashboard, utilizing Django backend for user authentication and game metrics processing.

Traffic Simulation to improve local traffic

• Developed an advanced traffic simulation system using Python and Unity, combining data analysis with 3D visualization to model urban traffic patterns and optimize flow efficiency in real-time.
• Implemented AI-driven vehicle behavior and pathfinding algorithms in C#, along with Matplotlib visualizations for traffic density analysis, resulting in actionable insights for reducing congestion and improving traffic management