Pradeep Mandal Email & Phone Number

United States

Santa Monica, California, United States

Throughout my career, I've demonstrated proficiency in leveraging Python and Pandas for efficient data collection and cleaning, ensuring the availability of high-quality datasets for analysis. My expertise extends to designing and conducting A/B tests using tools like Google Optimize, ensuring statistically sound experiments to validate data-driven decisions.I've utilized a combination of scikit-learn and TensorFlow to engineer relevant features, enhancing model performance and interpretability. This includes conducting extensive hyperparameter tuning and cross-validation to optimize model performance and generalization. Additionally, I've developed Python modules tailored for day-to-day business activities, incorporating machine learning and predictive analytics to streamline processes effectively.An area of specialization for me has been the utilization of advanced statistical models such as the Gamma-Gamma model as an extension to the Pareto model, focusing on purchase count, lifetime, and monetary value. I've also leveraged Python, PyTorch, and PySpark to design and implement advanced machine learning models, enhancing data-driven decision-making processes across various projects.I've demonstrated proficiency in cloud computing platforms such as MS Azure, harnessing cloud resources for scalable and efficient data processing and machine learning tasks. Additionally, I've managed data using Apache Cassandra, employed SAS for extracting actionable insights from complex datasets, and optimized distributed data processing using Apache Spark and HDFS.Moreover, I've applied various forecasting models like ARIMA and Holt-Winter, implemented machine learning algorithms for customer churn analysis and sentiment analysis on customer feedback. I've also designed, built, and deployed machine learning models using Azure Machine Learning in a production-ready hosted environment, demonstrating my ability to deliver tangible solutions to complex business problems.

Louisville, Kentucky, United States

In my role as a Data Analyst, I collaborated closely with the Principal Data Scientist, employing statistical modeling and machine learning techniques to extract insights from procurement data. I developed customized dashboards showcasing operational trends and optimized their performance, reducing loading time by 20%.Data wrangling was a significant aspect of my responsibilities, involving cleaning and reshaping data using the pandas library. I used a diverse toolset including SQL, R, Java, Scala, Python, and Apache Spark for in-depth analysis, presenting actionable insights to both management and technical teams.I led the establishment of an Azure Cloud Enterprise Data Platform, seamlessly connecting resources such as ADF, Databricks, and ADLS GEN2. I utilized machine learning algorithms like linear regression, Random Forests, and K-means to derive insights, implementing Azure DevOps for streamlined deployment.Working in the Hadoop/Hive environment, I used Spark/Scala, R, and Python for regex tasks, ensuring data integrity throughout the mining process. My expertise spans various machine learning models such as Logistic Regression and K-means clustering, applied to address analytical challenges effectively.Using R and Python, I conducted Exploratory Data Analysis to assess data effectiveness, facilitating informed decision-making. Lastly, I deployed statistical algorithms like Multivariate Regression and Random Forest models to identify and mitigate risks successfully.

Hyderabad, Telangana, India

As a Data Analyst specializing in Fraud and AML, I collaborated with clients to understand their challenges and integrated traditional transaction data with external sources using AWS EMR. I transformed SQL Server data to Hadoop Clusters, conducted data cleansing and feature engineering with Python's NumPy and Pandas.Implemented algorithms like Naive Bayes, Logistic Regression, SVM, Random Forest, and gradient boosting using Scikit-learn. Collaborated with ETL teams for data uniformity and used advanced SQL queries for cleansing. Conducted exploratory data analysis on different diagnoses datasets.Developed a comprehensive data preprocessing pipeline with Python's pandas, implemented multi-layer Neural Networks using TensorFlow and Spark. Conducted behavioral modeling and customer segmentation with K-means Clustering.Managed models using Oozie, updated Fraud predictions in AWS S3, and tested AML model performance for detection rates and false positives. Created reports and dashboards with Tableau for both technical and business audiences. Utilized GitHub for version control and collaboration with teams.

Mumbai, Maharashtra, India

In my role, I played a vital part in identifying key variables and conducting necessary data cleaning, manipulation, and preparation. Throughout the research process, I actively participated in all phases, including data collection, cleaning, mining, model development, and visualization.Collaboration with data engineers and operational teams was integral to collecting internal system data that met analytical requirements. Leveraging AWS cloud infrastructure, I contributed to scalable and cost-effective data solutions, utilizing its capabilities for storage, processing, and analysis.Using SQL queries, I refined attributes, relationships, and eliminated unwanted tables/columns. Additionally, I employed Spark SQL API in PySpark and utilized a C# connector to execute SQL queries, facilitating efficient data extraction and loading.For data imputation, I utilized the Scikit-learn package in Python, while data processing tasks were accomplished using Python libraries like NumPy and Pandas. In R, I employed the ggplot2 library for data visualizations, enhancing our understanding of customer behaviors.Statistical modeling was a key aspect of my work, where I implemented XGBoost machine learning software package in R to determine predicted probabilities for each model, contributing valuable insights to our analysis.