Looks like you have JavaScript disabled. For the full Twine experience, you will need to re-enable it.

I’m Varsha Reddy Madireddy, an AI/ML and Generative AI Engineer with 5+ years of experience building production-grade solutions across LLM applications, NLP, RAG, and agentic workflows. I focus on turning prototypes into reliable systems by combining Python, FastAPI, SQL, and modern AI frameworks with careful prompt engineering, structured outputs, and tool/function calling. I’ve worked on enterprise RAG and intelligent search using Azure AI Foundry and Azure AI Search, implementing hybrid retrieval, ranking/reranking, metadata filtering, and citation-based responses. I also build and evaluate AI agents using LangChain/LangGraph, AutoGen, Claude Agent SDK, and MCP, with strong emphasis on observability (LangFuse), quality evaluation (golden datasets and LLM-as-a-judge), and responsible AI/security controls to improve latency, reliability, and cost.…I’m Varsha Reddy Madireddy, an AI/ML and Generative AI Engineer with 5+ years of experience building production-grade solutions across LLM applications, NLP, RAG, and agentic workflows. I focus on turning prototypes into reliable systems by combining Python, FastAPI, SQL, and modern AI frameworks with careful prompt engineering, structured outputs, and tool/function calling. I’ve worked on enterprise RAG and intelligent search using Azure AI Foundry and Azure AI Search, implementing hybrid retrieval, ranking/reranking, metadata filtering, and citation-based responses. I also build and evaluate AI agents using LangChain/LangGraph, AutoGen, Claude Agent SDK, and MCP, with strong emphasis on observability (LangFuse), quality evaluation (golden datasets and LLM-as-a-judge), and responsible AI/security controls to improve latency, reliability, and cost.

Varsha Reddy Madireddy

Data Analyst, AI Engineer, Data Scientist, +3





I’m Varsha Reddy Madireddy, an AI/ML and Generative AI Engineer with 5+ years of experience building production-grade solutions across LLM applications, NLP, RAG, and agentic workflows. I focus on turning prototypes into reliable systems by combining Python, FastAPI, SQL, and modern AI frameworks with careful prompt engineering, structured outputs, and tool/function calling. I’ve worked on enterprise RAG and intelligent search using Azure AI Foundry and Azure AI Search, implementing hybrid retrieval, ranking/reranking, metadata filtering, and citation-based responses. I also build and evaluate AI agents using LangChain/LangGraph, AutoGen, Claude Agent SDK, and MCP, with strong emphasis on observability (LangFuse), quality evaluation (golden datasets and LLM-as-a-judge), and responsible AI/security controls to improve latency, reliability, and cost.…I’m Varsha Reddy Madireddy, an AI/ML and Generative AI Engineer with 5+ years of experience building production-grade solutions across LLM applications, NLP, RAG, and agentic workflows. I focus on turning prototypes into reliable systems by combining Python, FastAPI, SQL, and modern AI frameworks with careful prompt engineering, structured outputs, and tool/function calling. I’ve worked on enterprise RAG and intelligent search using Azure AI Foundry and Azure AI Search, implementing hybrid retrieval, ranking/reranking, metadata filtering, and citation-based responses. I also build and evaluate AI agents using LangChain/LangGraph, AutoGen, Claude Agent SDK, and MCP, with strong emphasis on observability (LangFuse), quality evaluation (golden datasets and LLM-as-a-judge), and responsible AI/security controls to improve latency, reliability, and cost.

Available to hire

I’m Varsha Reddy Madireddy, an AI/ML and Generative AI Engineer with 5+ years of experience building production-grade solutions across LLM applications, NLP, RAG, and agentic workflows. I focus on turning prototypes into reliable systems by combining Python, FastAPI, SQL, and modern AI frameworks with careful prompt engineering, structured outputs, and tool/function calling.

I’ve worked on enterprise RAG and intelligent search using Azure AI Foundry and Azure AI Search, implementing hybrid retrieval, ranking/reranking, metadata filtering, and citation-based responses. I also build and evaluate AI agents using LangChain/LangGraph, AutoGen, Claude Agent SDK, and MCP, with strong emphasis on observability (LangFuse), quality evaluation (golden datasets and LLM-as-a-judge), and responsible AI/security controls to improve latency, reliability, and cost.

Skills

Experience Level

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Intermediate

Language

English

Fluent

Work Experience

AI Engineer / Generative AI Engineer at Morgan Stanley

August 1, 2024 - Present

Developed and orchestrated production-grade AI agent workflows using Python, LangChain, LangGraph, AutoGen, Claude Agent SDK, and MCP to automate enterprise processes. Built Generative AI solutions on Amazon Bedrock, integrating foundation models with enterprise RAG, vector search, and secure application services. Implemented RAG pipelines with Azure AI Foundry and Azure AI Search, including document chunking, embeddings, metadata enrichment, hybrid retrieval (vector/semantic + keyword/BM25), metadata filtering, and reranking to improve answer quality. Built Databricks-based ETL/ML workflows using PySpark and Delta Lake, and integrated model tracking via MLflow. Added Bedrock foundation model integration with prompt engineering, tool/function calling, guardrails, structured outputs, and security controls. Evaluated and optimized retrieval/response quality with LangFuse, golden datasets, LLM-as-a-judge, and retrieval/response metrics; monitored latency, token usage, and cost. Implemente

Data Analyst / AI Engineer at Morgan Stanley

August 1, 2024 - Present

Architected and implemented end-to-end machine learning pipelines, integrating data ingestion, preprocessing, feature engineering, model training, validation, and deployment to enable scalable predictive analytics. Designed and deployed enterprise LLM-based solutions using Retrieval-Augmented Generation (RAG) with vector databases and prompt engineering to automate knowledge retrieval across large unstructured datasets. Engineered high-performance data pipelines with Apache Spark and Apache Airflow to orchestrate transformations on large-scale data while ensuring data integrity and fault tolerance. Optimized complex SQL queries and data warehouse schemas with indexing, partitioning, and refactoring to boost data retrieval speed and dashboard performance. Deployed models on AWS SageMaker with automated retraining, version control, and monitoring to sustain performance in production. Implemented robust MLOps using Docker, Kubernetes, and CI/CD pipelines for seamless deployment. Built int

Machine Learning Engineer at Optum, India

January 1, 2023 - December 31, 2023

Built end-to-end ML pipelines using Python, scikit-learn, Pandas, NumPy, and SQL, reducing model development/data-prep effort by 30%. Developed predictive analytics models for classification, regression, and clustering, performing EDA and feature engineering to improve performance. Optimized scikit-learn and XGBoost models with feature engineering, cross-validation, and hyperparameter tuning, improving predictive performance by 25%. Implemented NLP pipelines for text classification, NER, categorization, and information extraction using SpaCy, NLTK, and Hugging Face transformers. Developed FastAPI REST APIs to expose trained models for enterprise integration. Implemented MLOps with MLflow, Docker, and CI/CD to improve reproducibility and reduce deployment/model-management effort by 35%. Monitored deployed model performance and data quality for drift; collaborated with data engineers, data scientists, and stakeholders; produced documentation and performance reports.

Machine Learning Engineer at Optum

January 1, 2023 - December 31, 2023

Built end-to-end ML pipelines using Python, Scikit-learn, Pandas, NumPy, and SQL, including data exploration, preprocessing, and feature engineering. Developed predictive analytics models (classification, regression, clustering) and improved model performance through feature engineering, cross-validation, and hyperparameter tuning. Created NLP pipelines for unstructured text using SpaCy, NLTK, and Hugging Face models for text classification, NER, categorization, and information extraction. Exposed models via REST APIs using FastAPI. Implemented MLOps practices with MLflow, Docker, and CI/CD to improve deployment reproducibility and reduce operational effort; monitored performance and data quality for drift.

Data Analyst / Machine Learning Engineer at Optum

January 1, 2023 - December 1, 2023

Developed and deployed machine learning models for classification, regression, and clustering, including advanced feature engineering and hyperparameter tuning to improve predictive accuracy. Performed statistical analyses and hypothesis testing to validate business drivers and inform decision making. Built robust ETL pipelines with Apache Airflow for reliable ingestion, transformation, and loading from multiple sources. Created interactive dashboards in Tableau and Power BI to provide real-time visibility into key metrics. Automated data processing and reporting via Python and SQL scripts, reducing manual workload. Conducted customer churn analysis and predictive modeling to drive targeted retention strategies. Optimized data processing on Snowflake and Google BigQuery, improving query performance and reducing costs. Developed time-series forecasting models to predict demand and guide resource planning. Collaborated with stakeholders to define KPIs and align analytics initiatives with

Data Analyst at Aakruti Software Solutions Pvt Ltd

September 1, 2021 - December 1, 2022

Performed data extraction, transformation, and analysis using SQL and Python to support reporting and data-driven decision-making. Built interactive dashboards with Power BI to track key metrics and performance indicators. Automated reporting workflows and data processing to improve efficiency. Conducted EDA to identify trends and anomalies, supporting operational improvements. Optimized SQL queries and designed relational schemas for scalable data storage. Assisted in developing machine learning models for predictive analytics and supported data migrations. Collaborated with stakeholders to gather requirements and translate them into analytical solutions. Participated in Agile sprints and documented data pipelines and workflows.

Python Developer / Machine Learning Engineer at Aakruti Software Solutions Pvt Ltd, India

September 1, 2020 - December 31, 2022

Developed Python applications and data processing solutions supporting analytics, automation, and ML initiatives. Built ETL/data preprocessing pipelines for databases, CSVs, APIs, and external sources, reducing manual effort by 35%. Conducted exploratory data analysis with Pandas and NumPy to identify anomalies and data quality issues. Trained and optimized scikit-learn models for classification, regression, and clustering, improving accuracy by 25% through feature engineering, selection, cross-validation, and hyperparameter tuning. Evaluated models using accuracy, precision/recall, F1-score, and ROC-AUC. Implemented NLP solutions for sentiment analysis, keyword extraction, and document categorization using NLTK and SpaCy. Built Flask/FastAPI REST APIs for extraction/validation/transformation/reporting and ML model integration, reducing repetitive manual processes by 40%. Integrated PostgreSQL/MySQL via SQLAlchemy, used Git/GitHub for version control, and containerized services with Do

Python Developer / Machine Learning Engineer at Aakruti Software Solutions Pvt Ltd

September 1, 2020 - December 31, 2022

Developed Python applications and data processing/ETL solutions for analytics, automation, and ML initiatives. Designed ETL pipelines from databases, CSVs, and APIs; performed data profiling/EDA with Pandas and NumPy; and built ML models using Scikit-learn for classification, regression, and clustering. Improved accuracy through feature engineering, selection, cross-validation, and tuning; evaluated models with standard metrics (precision/recall/F1/ROC-AUC). Built NLP solutions (text classification, sentiment analysis, keyword extraction, document categorization) using NLTK and SpaCy. Delivered REST APIs with Flask/FastAPI and integrated services with PostgreSQL/MySQL using SQLAlchemy. Containerized applications with Docker and used Git/GitHub for version control in an Agile environment.