Nandini K

About Nandini K

I’m an AI/ML Engineer with 4+ years of experience building and deploying end-to-end…

Experience

1. Ai Ml Engineer

   Capital One

   May 2025 — Present · US

   Architected and deployed a real-time fraud detection pipeline using Python, and AWS SageMaker, enabling streaming inference on transactional data that reduced fraudulent transaction latency by 27% and improved proactive blocking efficiency across payment gateways.•Implemented deep learning-based credit risk scoring models leveraging TensorFlow and XGBoost, achieving a 19% uplift in prediction accuracy through advanced feature engineering, cross-validation, and continuous retraining using version-controlled datasets.•Developed RAG-enabled NLP pipelines using Transformers (BERT, GPT-style LLMs) and NLTK to automate sentiment classification, information retrieval, and named entity recognition, reducing manual review time from 10 hours to under 40 minutes per batch.•Optimized distributed model training workflows using Apache Spark integrating Agentic AI frameworks for automated experiment management, hyperparameter tuning, and deployment cycles across multi-GPU clusters orchestrated via Kubernetes.•Created interactive visual analytics dashboards with Seaborn, and Pandas to monitor RAG model outputs, fraud detection drift, data quality metrics, and retraining schedules, ensuring regulatory compliance and actionable insights.•Engineered a serverless LLM deployment architecture on AWS Lambda, integrated with Docker containers and GitHub Actions, enabling low-latency, scalable inference requests for customer-facing microservices.•Implemented large-scale data ingestion and preprocessing pipelines using Apache Hadoop & Hive ensuring smooth ETL execution for over 3 TB of financial transactions daily, improving downstream accessibility for analytics and AI workloads.

Education

• Robert Morris University

  Master's degree, Data Analytics