CSE Student β’ Full-Stack Developer β’ AI Enthusiast β’ Real-Time Systems
Iβm a CSE student and developer passionate about building practical software with clean interfaces and meaningful user experiences.
My interests include full-stack development, AI-powered applications, real-time systems, and modern web technologies.
- π Building and improving full-stack applications
- π€ Exploring AI-powered products and intelligent systems
- π§© Working on projects that solve practical problems
- π Continuously learning new technologies and improving my development skills
{
"name": "Kodi Roshan",
"role": "CSE Student | Full-Stack Developer | AI Enthusiast",
"focus": [
"AI-powered applications",
"Real-time web applications",
"Machine learning projects",
"Modern web experiences"
],
"currently_working_on": [
"Better UI/UX",
"Real-world applications",
"Accessibility improvements",
"SEO and performance"
],
"tech_mindset": "Building, Learning, and Growing Every Day."
}π StudySync
AI-powered learning platform with document analysis, RAG-based Q&A, intelligent tutoring, and automated quiz generation.
Tech: Next.js β’ TypeScript β’ Prisma β’ PostgreSQL β’ Redis β’ Supabase β’ Qdrant
π§ DocFin AI
Multimodal document intelligence platform for PDF Q&A, classification, and risk analysis with grounded AI responses.
Tech: Next.js β’ TypeScript β’ Gemini β’ RAG β’ Qdrant β’ Supabase
π Relay
Collaborative media review platform with timestamped video comments and image annotations, powered by Google Drive integration.
Tech: Node.js β’ Supabase β’ Google Drive API β’ OAuth β’ RLS
π¬ SyncTalk
Real-time chat application with room-based messaging, persistent chat history, Web-Socket communication, REST APIs, and responsive UI.
Tech: React β’ Node.js β’ Express.js β’ Socket.IO β’ MongoDB
π§© SyncSolve
Full-stack puzzle platform featuring Sudoku solving and chess analysis through an interactive web interface.
Tech: Next.js β’ TypeScript β’ Tailwind CSS β’ Spring Boot
π Car Price Prediction
Machine learning web application for used-car price prediction using data preprocessing, feature engineering, and a trained regression model.
Tech: Python β’ Flask β’ Scikit-Learn
Full-Stack Development β’ Artificial Intelligence β’ Machine Learning β’ Real-Time Applications β’ UI/UX β’ Problem Solving β’ Cloud Technologies
Building Side Projects β’ Exploring New Technologies β’ Listening to Music β’ Playing Badminton β’ Playing Guitar β’ Learning New Tools
Portfolio β’ GitHub β’ LinkedIn
Building, Learning, and Growing Every Day π




