Mahesh Khumkar

I’m an AI & Data Science Engineer with hands-on internship experience building and deploying production AI systems. I’m proficient in Python and leading ML/DL tools like PyTorch, TensorFlow, and Scikit-learn, along with NLP and vision libraries such as spaCy and OpenCV. I focus on turning models into usable applications, not just experiments.

I’ve independently developed and deployed 2 live AI-powered applications. My strength is the full ML lifecycle—EDA, model training, SHAP explainability, Docker containerisation, REST API development, and cloud deployment on AWS and Render. I care about accuracy and interpretability, and I build systems that are maintainable end-to-end.

In my Data Science internship, I developed a real-time multimodal emotion detection system combining facial (OpenCV/CNN), voice (MFCC/PyTorch), and text (NLP/spaCy) inputs. I integrated Docker and deployment data management (PostgreSQL/Firebase), and I improved performance using ensemble fusion across modalities. I’m excited to keep building production-ready AI that delivers measurable user impact.