Midhun K

I’m a Computer Science undergraduate with experience in project development, UI/UX design, and web development. I build user-friendly digital experiences, and I enjoy turning ideas into working products that people can actually use.

In recent projects, I built a collaborative design platform that generates UI components from natural-language prompts using the Gemini API, with interactive rendering via Fabric.js. I also enabled real-time multi-user editing using a custom WebRTC + Y.js setup, and I developed an ML-based multiclass software defect prediction model using CodeBERT and AST (including fused attention), trained with NumPy/ScikitLearn/Seaborn/Matplotlib. During my internship at Intrainz Solutions, I developed fraud detection models using Scikit-learn and XGBoost, handling preprocessing, feature engineering, and class imbalance, then evaluated performance with accuracy, precision, recall, and confusion matrices. Outside of development, I’ve coordinated technical infrastructure and mentored students in UI/UX design and C, and I led education outreach for C programming through IEEE.