Anshu Gupta

About

Machine learning engineer with hands-on experience across computer vision and tabular ML — from a drowsiness-detection classifier (99.96% test accuracy, compressed 90% for edge deployment) to a ride-completion model that lifted ROC-AUC from 0.52 to 0.72. Comfortable across the full pipeline: EDA, feature engineering, model evaluation and validation (ROC-AUC, F1, mAP), and deployment (Docker, FastAPI, TensorFlow Lite). Also experienced with LLMs, RAG, and LangChain/LangGraph-based AI agents, with a strong foundation in Python and statistics.

Skills & Expertise (44)

Work Experience

Education

Certifications