Hire Govinda R. - Vector Database Engineer

* Developed predictive portfolio optimization models (XGBoost, LightGBM, PyTorch) integrating macroeconomic indicators, improving investment strategy accuracy by 18%. * Built a scalable time-series forecasting framework (ARIMA, Prophet) on AWS SageMaker for asset price predictions, reducing portfolio risk exposure. * Designed and deployed a real-time fraud detection pipeline using Autoencoders & Isolation Forest, integrated with Spark & Kafka for high-volume anomaly detection. * Automated data ingestion & transformation with Apache Airflow and AWS Glue, consolidating Bloomberg, Reuters, and internal trading data into a unified analytics platform. * Implemented an LLM-powered market sentiment analysis workflow using LangChain, FAISS, and OpenAI API to generate daily investment insights from analyst reports and news feeds. * Enhanced ESG scoring models through feature engineering, clustering, and NLP-driven classification on Azure ML Studio to support sustainable investment strategies.

* Engineered a real-time credit risk assessment system (Logistic Regression, Random Forest, XGBoost) on Databricks, reducing default prediction errors by 22%. * Built deep learning-based fraud detection models (LSTM, CNN in TensorFlow) for sub-200 ms transaction anomaly detection. * Created an automated A/B testing framework in Python to evaluate loan approval policies, improving data-driven decision-making. * Migrated risk analytics pipelines to AWS cloud-native architecture, improving scalability & processing time by 40%. * Integrated market volatility forecasting models (GARCH, Prophet) into trading strategy engines for better hedging performance. * Established secure MLOps pipelines with Docker, Kubernetes, and Jenkins for reproducible, compliant model deployments. * Partnered with quantitative analysts to optimize feature selection &hyperparameter tuning for large-scale financial datasets, increasing model stability & interpretability.