RAVI CREATIONS

I build ML and data science systems using Python, TensorFlow, PyTorch, and scikit-learn—especially predictive modeling, deep learning, and computer vision. In my M.Tech work, I fused a MobileNetV3-based feature extractor with real-time IoT sensor signals for plant disease detection and environmental advisory.

On the deployment side, I’ve delivered inference pipelines as FastAPI backends (with Hugging Face Spaces) and paired them with a frontend (Vercel), targeting sub-200ms latency plus offline support and edge inference on Jetson Nano. I also trained on the MANIT HPC cluster with PBS job scheduling and designed a late-fusion architecture for high-accuracy classification.

I’m continuing this through projects like CareGap AI, where I predict patient readmission risk with XGBoost and add SHAP-based explainability, plus an LLM-driven SMS/WhatsApp engagement engine with emergency escalation. I’ve also worked on ML-driven cybersecurity by building and evaluating intrusion detection models and integrating the T-Pot honeypot framework for proactive monitoring.