Prudhvi Raj D

Developed and refined Python-based quantitative algorithms for fixed-income and taxable SMA portfolios, modeling yield optimization, duration targeting, and after-tax performance to support systematic tax-advantaged investing. Built automated transition analysis pipelines simulating real-time transitions from client bond portfolios into AB-managed SMA models, integrating pricing, yield, and tax impact analytics via REST APIs and SQL backends. Created automated proposal generation frameworks that produced optimized model portfolios for cash-funded or partial-transition accounts, incorporating liquidity constraints, benchmark alignment, and tax efficiency scoring. Engineered trade rationale algorithms that automatically explained trading activity, yield pickup, and tax-loss harvesting benefits—enabling transparent reporting and advisor insights via Service Center dashboards. Developed transition scorecards and onboarding analytics summarizing tax alpha, realized gains/losses, and portfolio changes, automating report delivery to portfolio managers and advisors. Implemented optimization and simulation models in Python using SciPy, NumPy, and PyPortfolioOpt to maximize after-tax yield while controlling for credit risk, duration drift, and benchmark tracking error. Designed and deployed quant analytics APIs and microservices (Flask/FastAPI, .NET) integrated with Service Center and internal platforms, supporting end-to-end automation of portfolio rebalancing and optimization.

Built and deployed scalable ML models using TensorFlow, PyTorch, and Dask for claims risk scoring and healthcare prediction. Integrated them into enterprise risk platforms, improving personalization accuracy and reducing healthcare costs while ensuring high availability in production systems. Developed advanced fraud detection and chronic disease prediction pipelines using LSTMs, GRUs, XGBoost, and Random Forest. Enhanced automated risk assessments and care coordination across Medicare and Medicaid services, enabling better preventive care and fraud mitigation. Designed time-series forecasting solutions for healthcare utilization and claims prediction using historical data. Improved demand planning accuracy by over 30%, optimizing resource allocation, reducing clinical staffing strain, and enabling proactive operational decision-making. Built NLP models leveraging LSTM and early Transformer techniques (BERT) for processing call center transcripts, enhancing compliance monitoring and service quality. Designed ETL pipelines for text data preparation, integrated with SQL-based data warehouses for analytics and reporting.Delivered topic modeling, sentiment analysis, and entity recognition to generate insights for customer experience improvement and compliance reporting

Designed and deployed multi-agent AI workflows using CrewAI and MCP, orchestrating autonomous reasoning, retrieval, and tool execution. Streamlined enterprise marketing operations while enabling agents to collaborate dynamically. Delivered end-to-end automation with high efficiency. Developed prompt engineering frameworks with reusable templates, dynamic context management, and version control. Standardized interactions across multiple LLM applications, ensuring consistency and high-quality outputs. Improved response relevance and reduced iteration time. Implemented retrieval-augmented generation (RAG) pipelines with optimized chunking, embedding strategies, and vector databases (Pinecone, FAISS, ChromaDB). Enhanced contextual understanding of queries and generated highly accurate answers for internal stakeholders. Designed and deployed Transformer-based NLP models (BERT, RoBERTa) for analyzing call/chat transcripts, enabling automated compliance checks and customer coaching insights. Leveraged Databricks and Spark to design end-to-end ML pipelines, including feature engineering, predictive modeling, and time series forecasting for large-scale datasets. Integrated AI/ML workflows with cloud platforms (Azure/AWS/GCP) and developed data-driven optimization solutions for business operations

Collaborated with business teams to gather requirements, design meaningful data models, and translate functional needs into analytics-ready datasets. Enhanced data standardization practices and improved documentation for long-term reporting reliability. Partnered with data engineers to extract and process historical and real-time data using Hadoop tools such as Hive, Sqoop, Pig, and MapReduce. Ensured consistency, timeliness, and accuracy in ingestion workflows across enterprise-scale datasets. Designed and optimized Hive UDFs and advanced SQL queries to handle complex data types, perform joins, aggregations, and transformations. Delivered efficient pipelines that enabled downstream reporting and analytics with improved performance. Built clustering and segmentation models using K-Means and hierarchical algorithms in Python and R. Identified behavior patterns, grouped similar customer segments, and provided actionable insights for targeted marketing and churn reduction. Developed time-series forecasting models with ARIMA and exponential smoothing in R to predict demand trends and operational metrics. Supported supply chain and planning teams with forecasts that reduced uncertainty and resource strain. Cleaned, transformed, and profiled raw data using Pandas, NumPy, and dplyr, ensuring consistency and data quality across analytical workflows. Implemented validation scripts to catch anomalies early and improve reliability of downstream reports. Created interactive Tableau dashboards with custom charts, maps, and trend visualizations to monitor KPIs and business metrics. Published secure dashboards to Tableau Server, empowering executives and teams with timely, self-service analytics.