Anshu User

I’m a data analyst and machine learning practitioner building end-to-end solutions that convert complex datasets into measurable impact. In my JEH Aerospace internship, I designed and deployed Power BI dashboards that cut report generation time by 20% for senior stakeholders.

I also focus on operational decision-making through ML-driven insights. I improved on-time maintenance scheduling by 15% across 2 business units by extracting insights from structured sensor data using Python and SQL, and I reduced ad hoc data request tickets by 30% by creating a unified data catalog across 12+ sources.

Beyond analytics, I ship production-ready models and systems. I delivered an NLP cyberbullying classifier achieving 89% accuracy across 6 toxicity categories, and I built a distributed task queue system in Java with Spring Boot and Redis that supports 10,000+ tasks/min, reduced latency by 43%, and achieved 99.7% uptime in load tests. I’m equally motivated by applied deep learning work—ranging from a 3D CNN multimodal Alzheimer’s model (91.3% accuracy, 0.94 ROC-AUC) to construction-site detection and segmentation for real-world safety.