Sri Haran

I've built AI and full-stack applications using Python, Flask, React.js, Node.js, and MongoDB, including a Heart Disease Prediction platform that combines CNNs, SHAP explainability, and federated learning.

For the heart disease project, I trained a CNN on 1,500+ ECG images, achieving 87% AUC and 84% F1-score. I also implemented federated learning across three simulated hospital nodes with less than 2% accuracy loss while helping preserve patient privacy.

During my internship at Asscentz Company, I developed REST APIs and responsive React modules, reducing API response time by about 25% and raising mobile usability from 68 to 89. I'm available for immediate joining and enjoy building practical LLM, RAG, agentic AI, and cloud-deployed applications.