Piyush Rawat

I’m a Machine Learning Intern focused on building production-minded ML solutions—trend prediction, anomaly detection, and real-time monitoring—while keeping pipelines reliable and repeatable. At Internselite, I designed ML models for trend prediction and anomaly detection using XGBoost, Isolation Forest, and Autoencoders, and built Airflow pipelines with data validation, drift checks, and scheduled retraining that reduced manual analytics work by 70%.

I also develop end-to-end ML pipelines and deployment-ready services. At Coincent, I created preprocessing + feature engineering workflows, added hyperparameter tuning with Scikit-learn and TensorFlow, set up MLflow experiment tracking, and developed FastAPI endpoints with Dockerized components—improving prediction accuracy by 18-25% and cutting manual testing time by 40%. My projects include a real-time AI-Driven Network Monitor with DDoS detection (93% detection accuracy) and an AI File Organizer (92% accuracy) that reduced manual sorting time by 80%.