Pratik Kumar Email & Phone Number

Jersey City, Nj, Us

Jersey City, Nj, Us

Northbrook, Il, Us

• Developed various logistic and linear regression models for fraud detection, target marketing and churn prediction in R• Extract, Transform and load data from sources systems to AWS Data Storage services using a combination of AWS Redshift, AWS S3, AWS EMR, AWS Glue and AWS Lambda.• Created a prediction model in R to identify the clime submission pattern • Developed storytelling dashboards in Tableau Desktop by taking various analytical reports from multiple data sources• Worked on AWS Data Bricks workspace, Cluster , Mount , Secret Key and Delta DB setup. Using Data Bricks to Store Schema on External SQL Server.• Finalizing the requirements and analyzing the In-scope and Out-scope for requirements• Discussion on scope for implementation of Analytical models with Business users• Collecting and transferring all data into Data Lake and use of Amazon Redshift or machine learning, network data transfer to cloud.• Writing SQL queries for performing CRUD operations on data from DB Server• Perform Data Analytics using Excel, SQL and R including NoSQL databases using Cassandra, PostgreSQL ,Mongo DB and Azure Cosmos DB

• Analyzed large customer datasets of different demographic groups to gain insight into dining tendencies using Map Reduce in Java and subsequently supporting the decision of running right campaign leading to much-improved redemption rate• Used AWS stack for machine learning, network data transfer to cloud.• Using AWS Glue orchestrated jobs , Read / Write data from multiple sources and destinations , performed data transformed and stored on destination.• Performed the deep learning on the dataset using PyTorch in TorchServe environment. • Used Random forest algorithm to help identify loyal customers and predicted the likelihood of customers buying a recommended product

• Performed Logistic Regression, Classification, Random Forests and Clustering in Python.• Worked on Linear Discriminant analysis, Greedy Forward Selection, Greedy Backward Selection and Feature Selection, Feature reduction algorithms like Principal Component Analysis and Factor Analysis.• Used multiple machine learning algorithms, including random forest and boosted tree, SVM, SGD, neural network, and deep learning using TensorFlow.• Skilled with Python parsing, manipulating, and converting data to and from a wide range of formats (CSV, json, XML, html, etc).• Familiarity with the AWS ecosystem, including tradeoffs between services that meet overlapping needs.• Responsible for elevating the code into CI/CD process using development, test, and Production environments on schedule. Provides Production support when needed• Involved in Building the CI/CD Pipelines and automating the pipeplines• Developed data pipeline using Flume, Sqoop, Pig and Java MapReduce to ingest customer behavioral data and financial histories into HDFS for analysis. • Developing Spark applications using Spark - SQL in Databricks for data extraction, transformation and aggregation from multiple file formats for analyzing & transforming the data to uncover insights into the customer usage patterns.• Using Informatica PowerCenter Designer analysed the source data to Extract & Transform from various source systems(oracle 10g,DB2,SQL server and flat files) by incorporating business rules.• Created customized BDM mappings/workflows for incremental using Informatica DEI and deploy them as part of an application on Data Integration service.

Responsibilities• Performed data extraction, aggregation, log analysis on real time data using Spark Streaming• Experience working in project with data visualization, R and Python development, Unix, SQL• Performed exploratory data analysis using numpy, matplotlib and pandas• Expertise in quantitative analysis, data mining, and the presentation of data to see beyond the numbers and understand trends and insights.• Experience in using the Lambda functions like filter (), map () and reduce () with pandas Data Frame and perform various operations.• Implemented Principal Component Analysis and Liner Discriminate Analysis.• Used Pandas API for analyzing time series. Creating regression test framework for new code.• Creating complex SQL queries and scripts to extract and aggregate data to validate the accuracy of the data and Business requirement gathering and translating them into clear and concise specifications and queries.