Kishore Kannapurakkaran

This Data Lakehouse is a centralized data hub based on the Azure cloud platform to facilitate data for multiple data marts and provide data feeds for external vendors and internal applications (APIs). I designed the ELT framework to fetch data from on-premise systems into the data lake, apply business rules, and load it into Azure SQL database tables. Also, I created a source to target a mapping document with detailed transformation rules.

This application enables schools' technical support teams to access crucial consolidated tech information for each school based on the Microsoft Dynamics 365 CRM platform. I designed and developed an ETL change capture process using IBM DataStage to update a data repository using SQL Server, which stores datasets to be uploaded into CRM. I also developed data transformations using Scribe Insight to upload the daily updates into Dynamics CRM 365.

I built a data integration (DI) layer using IBM DataStage, which fed data into the SAP FS-CD system to replace existing general ledger (GL), finance and administration (FA), accounts receivable (AR), premium processing, collection and disbursement, and advanced placement (AD) applications. The data is sourced from a legacy IBM AS/400 system. The DI interfaces include data for a lender, borrower, loan certificate, premiums, claims, refunds, and contract underwriting.

I converted around 2,500 ETL jobs and sequences from IBM InfoSphere DataStage to an open source tool Pentaho Data Integration (PDI). Additionally, I converted complex ETL jobs as well as the logging, auditing, and error handling components from IBM InfoSphere DataStage into PDI transformations and jobs.

I integrated AutoSupport log data from the install base system with the enterprise business intelligence (BI) analytics to provide visibility to active installed products' current state of operation and configuration. It gives a much greater depth into client systems' workings and customer behavior and better identifies cross-selling, up-selling, and tech-refresh opportunities.

I created an analytical reporting mart to track and trend the state of the business, i.e., identify trends in total assets under management, the number of accounts, plan types, cash flow, business line growth, plan movement, and more. I designed and developed ETL batch jobs in IBM DataStage to perform a daily change data capture and load Type 1 and Type 2, slowly changing dimensions and fact tables.

This is a retail data warehouse re-architecture project of a legacy system to a hub-spoke architecture. This project's goals were:

• Reducing cost and time of delivery for the data warehouse development.
• Automating a decision science process and responding faster to new decision science requests.
• Using predictive models to drive the service model and show measurable results.
• Filling data gaps in the data warehouse to support strategic questions involving the web, call center, mail, and email.
• Improving the campaign targeting, selection, and measurement.