Based on the subscription and payment plan type users, data was seggregated as prepaid and postpaid. I have architected and wrote a python module to process the entire data . So that this tailored data can be run through some analytical models and run offering campaigns based on the age, head of the house, payment historry , zip code etc. Everyweek this data needs to be run through python model severl times and provide the output analytical platform. All this process is also automated and streamline in the last phase of the project.

Our company has proprietary data analytics platform . Custom models are trained and injected into the platform depending on workflow of the data. I have written java modules to tailor the data and train the models based on the data. Resulted outputs will be used by in-house data analysts/scientist and adjust the signals for benching marking the false positives on the model.

General Motors : Sales Incentive Process fraud application, Bank Manage Systems

FICO : Financial and Debit Card Fraud Integrations and development

Opera Solutions : Mobile Data Analytics Platforms, Threat Analysis Platform

Zappos : Email Markerting and Supporting Applications

Architected new solutions and maintained the entire Zappos email marketing enclave in AWS cloud. This includes triggered and broadcast emails. Implemented transactional emails , account management emails for new speciality goods sites with in zappos under AWS master account. Maintained IAM roles and policies for the entire team. Created and maintained datapipelines from between , RDS, Redshift and S3 buckets. Created and maintained glue crawlers to load data from s3 to DB. Extensively worked on DynamoDB.

This project was developed for GM Incentive Audit Review Group to track and report all the sales audits done by the field Area Sales Managers. Application will detect the fraudulent incentives and bonuses used by the Dealers with the built-in intelligence algorithms. Application generates a PDF reports dynamically on the fly.

This project was developed for GM Treasury to track transaction fees associated with different banking transactions. BMS (Bank Management System) is an n-tier web application based on J2EE application architecture, Oracle 8i is used as database and Crystal Reports is used for reporting. BMS (Bank Management System) is deployed globally to a user base of 300 financial analysts.