Sai Karan Reddy User

I’m a Data Scientist with 5+ years of experience delivering production-ready machine learning, deep learning, and generative AI (LLM/RAG) solutions across financial services and large-scale systems. I focus on predictive modeling, anomaly detection, credit risk, and AML analytics—always with reliable deployment, monitoring, and governance in mind.

In my current role, I built and deployed predictive models for GPU/CPU failure prediction using LightGBM and PySpark, working with telemetry from 18K+ production nodes. I’ve engineered scalable time-series features, implemented drift monitoring and model governance with MLflow + PSI integrated into an Airflow scoring pipeline, and used SHAP-based explainability to make models interpretable and production-ready. I’ve also fine-tuned LLMs with LoRA/PEFT for hardware failure logs and deployed RAG pipelines with LangChain and FAISS to reduce diagnosis time.

Previously at JPMorgan Chase, I developed credit risk models scoring 1.5M+ loan applications while supporting RBI model governance, and I built AML anomaly detection systems that reduced false positives by 40%. I engineered large-scale graph features for fraud network discovery and deployed scalable pipelines on AWS with real-time inference and batch scoring across millions of customers monthly. Across both roles, I combine strong experimentation with practical MLOps to create systems that stay trustworthy in production.