Data Scientist - AWS, ML, NLP, DL Specialist

Varun Kumar Raju Addepalli

Data Scientist

******************@*****.*** 240-***-**** LinkedIn

SUMMARY

● Data Scientist with 5 years of experience leveraging Python, Machine Learning, Deep Learning, Amazon Web Services (AWS), SQL, and Data Management to extract insights from complex datasets.

● Skilled in utilizing AWS cloud services like S3, Redshift, and Lambda to build and maintain robust data pipelines and effectively communicate insights through Tableau and Power BI data visualizations.

● Ability to apply statistical methods, such as Hypothesis Testing, and machine learning algorithms, including Decision Trees, to build and evaluate data models, conduct A/B testing, and effectively communicate insights through compelling data visualizations.

● Proficient in deep learning frameworks like convolutional Neural Networks (CNN) used in computer vision tasks, Recurrent Neural Networks (RNN), LSTM, and Transformer architecture used in Natural language (NLP) processing.

● Expertise in leveraging a wide range of Python libraries, including Matplotlib, Scikit-learn, Seaborn, TensorFlow, Keras, NLTK, XGBoost, and PyTorch, for advanced data analysis, machine learning model development, and effective collaboration using Git. EDUCATION

Masters in Computer Technology

Eastern Illinois University, Charleston, IL

SKILLS

Languages: Python, R, SQL

IDE’s: Jupyter Notebook, Google Colab

Machine Learning: Linear, Logistic Regression, Decision Trees, Random Forests, Naive Bayes, SVM Deep Learning: CNN, ANN, RNN, LSTM, LangChain, Hugging Face Transformers (BERT, GPT-3), Natural Language Processing, Large Language Model Cloud/Visualizations: AWS (S3, RedShift, Lambda, Glue, SageMaker), Tableau, Power BI, Excel, Looker Statistical Techniques: Hypothesis Testing, Decision Trees, Data Visualization, Data Modelling, A/B testing, Model Evaluation Packages and Frameworks: NumPy, Pandas, Matplotlib, Scikit-learn, Seaborn, TensorFlow, Git, Keras, NLTK, XGBoost, PyTorch Database: MySQL, PostgreSQL, MongoDB, SQL Server

PROFESSIONAL EXPERIENCE

CitiGroup, IL January 2024 – Present

Data Scientist

● Built and deployed machine learning models for credit risk assessment, utilizing techniques like logistic regression and decision tree algorithms, resulting in a 20% reduction in loan defaults and increased profitability.

● Leveraged the Linear platform to develop and deploy an innovative machine-learning model for anomaly detection in financial transactions, resulting in a 15% improvement in fraud detection accuracy.

● Developed an innovative image classification model utilizing a Convolutional Neural Network (CNN) for fraudulent document detection, achieving a 25% accuracy in identifying fraudulent checks. This significantly improved fraud detection rates and reduced financial losses.

● Established and deployed a LangChain-powered application for customer inquiries regarding investment options, automating initial query handling and significantly improving response times, resulting in a 15% reduction in customer service resolution times.

● Created a Random Forest model to predict customer support ticket priority with an F1-score of 0.92, resulting in a 30% reduction in average ticket resolution time and improved customer satisfaction by 12%.

● Utilized AWS S3 to build a scalable data lake, enabling the storage and processing of massive datasets for advanced analytics and facilitated the development of more sophisticated machine learning models, leading to a 20% increase in the accuracy of our predictive models.

● Conducted A/B testing on investment product landing pages, resulting in a 15% increase in conversion rates by identifying and implementing the most effective variations. This led to a significant increase in customer acquisitions and revenue generation. Tredence India June 2020 – November 2022

Data Scientist

● Deployed an ANN for anomaly detection in sensor data, achieving a 20% detection rate demonstrates expertise in developing and implementing deep learning solutions for complex real-world problems.

● Established advanced forecasting models ARIMA and XGBoost to improve accuracy by 15%, enabling data-driven decisions that drove revenue growth trends visualized in AWS Quicksight.

● Implemented a scalable data warehouse on AWS Redshift, enabling the analysis of massive datasets with minimal performance degradation to support the growth of our data-driven initiatives, leading to a 15% increase in data analysis efficiency.

● Built a robust recommendation system utilizing NLP techniques and NLTK library to enhance product discovery. Extracted meaningful insights from product descriptions and user reviews to create personalized recommendations, resulting in a 20% increase in click-through rates and a 30% boost in sales.

● Orchestrated the development of real-time Tableau dashboards showcasing trend analysis and vital metrics, enabling stakeholders to make data-driven decisions resulting in a 20% increase in revenue and a 15% decrease in operational risks

● Leveraged hypothesis testing techniques, including t-tests, to analyze campaign performance and identify key factors influencing campaign success, approach resulted in a 12% increase in click-through rates and improved campaign ROI. Fractal Analytic MH, India January 2019 – May 2020 Data Analyst

● Utilized AWS cloud services to develop and implement scalable data pipelines, enabling the analysis of large and complex datasets. This supported the growth of data-driven initiatives and improved the efficiency of data analysis processes by 20%.

● Leveraged cloud-based tools, including AWS Cloud, to build robust and scalable data pipelines that support critical operations reducing costs by 20% and enabling faster access to high-quality data for timely decision-making.

● Crafted and maintained interactive dashboards in Power BI, streamlining data analysis and reporting processes, resulting in a 20% reduction in report generation time and improved team efficiency.

● Streamlined scikit-learn to machine learning workflow, enabling faster model development and deployment, resulting in a 15% reduction in model development time and improved overall project efficiency.

● Ensured data quality and integrity by implementing data validation and cleaning routines within the PostgreSQL database, leading to a 15% reduction in data errors and improved data accuracy for downstream analyses.

● Developed and deployed an innovative TensorFlow-based model for customer churn prediction, achieving an accuracy of 25% model enabled proactive customer retention efforts, leading to significant cost savings and improved customer loyalty.

Contact this candidate