Goutham Chimata

• Architected and deployed a production-grade Retrieval-Augmented Generation (RAG) platform, improving analyst response times by 55%. • Designed a multi-agent orchestration system that automated workflows, reducing incident resolution time by 40%. • Established an ML observability framework, decreasing model issue detection time from 48 hours to 15 minutes across 200+ models.

• Developed a high-precision fraud detection system using PyTorch, TensorFlow, XGBoost, and Autoencoders, achieving 98% ROC-AUC in fraud classification. • Implemented feature engineering, PCA-based dimensionality reduction, and hyperparameter tuning to reduce false positives in banking transactions. • Deployed deep learning models on AWS using SageMaker, Lambda, and API Gateway for real-time fraud monitoring with sub-100ms latency.

Designed and deployed scalable machine learning solutions using Python, Scikit-learn, Pandas, and PySpark to process large-scale healthcare and operational datasets across enterprise platforms. • Built end-to-end ML pipelines for data ingestion, preprocessing, feature engineering, model training, validation, and deployment in cloud-native environments. • Developed predictive analytics models using Random Forest, XGBoost, and Neural Networks to improve operational efficiency, anomaly detection, and business forecasting workflows. • Engineered real-time and batch data processing pipelines using Apache Spark, Kafka, and distributed data frameworks to support high-volume enterprise data systems. • Integrated Generative AI and LLM-powered workflows using LangChain, vector search concepts, and RAG architectures for intelligent document retrieval and workflow automation. • Designed RESTful APIs and microservices using Python and FastAPI/Flask to serve ML predictions and AI-driven business applications.