I'm Sanket Jagadale — a UI/UX Designer,MERN stack developer who turns requirements into production-ready apps with AI-integrated features, clean APIs, and real-world impact.
Projects Shipped
AI Integrations
LeetCode Solved
Figma UI Projects
I'm Sanket Jagadale, currently pursuing a B.Tech in Computer Science at Parul University, Vadodara, Gujarat. During my academic journey, I developed a strong interest in problem-solving and learning new technologies. I have built projects such as DSA Tracker, VITRA AI, and RepoGPT, which further strengthened my passion for UI/UX Design, web development and AI.
I consider myself a quick learner and a logical thinker. For example, while working on coding problems and projects, I have successfully debugged complex issues and improved code performance by optimizing memory usage and efficiency.
My short-term goal is to secure a Software Development Internship where I can apply my skills, learn from experienced engineers, and contribute to meaningful projects. My long-term goal is to build a successful career as a Software Engineer in a leading product-based company.
My interests lie at the intersection of UI/UX Design, web development, AI integration, and competitive programming. I have worked with the Gemini API to build intelligent code analysis tools and have integrated real-world AI features into production-ready applications.
I am actively seeking UI/UX Designer, and entry-level Software Engineer opportunities where I can contribute to impactful products, collaborate with talented teams, and continue growing as an engineer.
Technologies I use to build, ship, and scale.
Real-world projects — built to solve problems, not just to look good.
Problem: Students preparing for tech interviews had no centralized tool to track their DSA progress, get feedback on code quality, or organize problems by topic.
Solution: Built a full-stack DSA Tracker with an integrated Gemini AI code analysis engine. Users can log problems, paste their solutions, and receive instant AI feedback on time/space complexity, code correctness, and improvement tips.
Problem: Traditional AI assistants provide generic responses and lack long-term memory, personalization, and understanding of individual user behavior.
Solution: Built an AI-powered Digital Twin platform that learns from user interactions, habits, goals, and feedback to deliver personalized responses, adaptive behavior, and contextual assistance.
Problem: Online shoppers often spend significant time searching across multiple e-commerce platforms to find the best price, discounts, and deals for a product.
Solution: Built an AI-powered deal-finding agent that aggregates and compares real-time product prices across Amazon, Flipkart, Myntra, Meesho, Croma, and Snapdeal, helping users instantly discover the best deals from a single interface.
Problem: Developers spend significant time exploring large codebases, understanding project architecture, and locating relevant code across multiple files.
Solution: AI-powered codebase assistant that enables developers to chat with GitHub repositories using RAG, semantic search, and intelligent code analysis with file-level references.
Problem: Traditional games don't leverage computer vision for real-time gesture recognition.
Solution: Real-time hand gesture detection using OpenCV + cvzone to play RPS against AI via webcam — no clicks needed.
Consistent problem-solving — the foundation behind clean, optimized code.
Head to LeetCode Profile for full stats
Have an opportunity or just want to say hi? I reply within 24 hours.
Open to SDE internships, full-time roles, and freelance projects. Let's build something great together.