Nand Kishore Sagi

• Created data pipeline DAGS to migrate data from various MySQL databases, Elasticsearch and DynamoDB tables to Snowflake in airflow
• Prepared functional and technical documentation data for warehouses.
• Built Snowflake databases, warehouse, roles, schemas and table structures for ETL pipelines.
• Created and provisioned required EC2 instances, IAM roles, security groups to host Airflow.
• Identified and fixed incorrect data issues in QuickSight reporting
• Created data Snowflake shares to share data with External customers
• Cooperated fully with product owners and enterprise architects to understand requirements.
• Contributed to internal activities for overall process improvements, efficiencies and innovation.
• Prepared documentation and analytic reports, delivering summarized results, analysis and conclusions to stakeholders.
• Completed quality reviews for designs, codes, test plans and documentation methods.
• Mapped data between source systems and warehouses.

(1.) Decision Analytics and Reporting System

• Designed and created multi-tenant data warehousing solution in Snowflake by integrating data sources from Salesforce and Snowflake OLTP sources.
• Setup Airflow in local using Docker and Sequential executor to perform unit testing for the developed DAGs.
• ETL process was orchestrated using Airflow in AWS (AWS MWAA)
• Data transformation from Variant data to de-normalized columnar data was done using Snowflake Javascript procedures and these snowflake procedures would be called in individual Airflow tasks.
• Enabled GitLab CICD pipeline to push the code (DAG files and configuration files) from GitLab repository to S3 bucket.
• Created Python module to query the registration database for client specific encrypted configuration data and decrypt them using appropriate client specific KMS key.

(2.) Automated looker code deployment pipeline

• Created the GitLab CICD pipeline to push looker code from development repository to appropriate tenant specific looker repo.
• Used GitLab REST API's to fetch the development repository and perform client specific updates in the lookML files and replace the existing GitLab client specific branch and tags with the latest sprint end code.
• Integrate the client specific looker branch with client specific salesforce app.
• Enabled integration of Looker with Salesforce.

(3.) Automated production metrics extraction

• Created SQL queries in snowflake production to extract metrics for case processing.
• Enabled extraction of production metrics for each client after the client moved to production.
• Enabled auto-notification via email and custom schedules for metrics generation for each tenant.

(4.) Automated Salesforce Metadata ETL

• Created ETL process in python utility to extract ERD (Entity Relationship Details) from Salesforce using Salesforce REST API's.
• Extracted JSON data was parsed and loaded into Salesforce metadata tables in Snowflake. JSON output from the API responses were stored as Variant data in the tables.
• Created views in snowflake on top of the Salesforce metadata tables to help ensure consistency between the Data Model and physical implementation in Salesforce.

(5.) Automated Salesforce Metadata ETL

• Created ETL process in python utility to extract ERD (Entity Relationship Details) from Salesforce using Salesforce REST API's.
• Extracted JSON data was parsed and loaded into Salesforce metadata tables in Snowflake. JSON output from the API responses were stored as Variant data in the tables.
• Created views in snowflake on top of the Salesforce metadata tables to help ensure consistency between the Data Model and physical implementation in Salesforce.

(6.) Automated Test Data Seeding in Salesforce

• Identified and solved the need for test data generation to improve integration testing coverage
• Designed and created python based utility that would seed a salesforce org with multiple case data records. Each data record had multiple combinations of approximately 400 data points.
• A Excel file listed records with requisite data scenarios which is then read by a Python based utility.
• The utility would then create a Salesforce GraphAPI compatible JSON structure for each case data record.
• The multiple JSON files would then be loaded into the Target salesforce org by the utility using the Salesforce Graph API's.

(1.) Automated SDDM ETL

• Identified the need for publishing the data in SQL Developer Data Modeler (SDDM) XML files.
• Created a Python utility to process multiple XML files and load the content in Snowflake database.
• Created a GitLab CICD pipeline to publish the data model data to Snowflake as soon as the latest data model was pushed to GitLab.
• Created appropriate database, roles, schemas, tables, views in snowflake to help visualize the data model based on the data loaded.
• Followed SCD type 2 design for capturing the data model design.

(2.) Snowflake Deployment Utility

• Created a multi-tenant deployment utility in Python to automate the infrastructure setup, deployment and testing of Snowflake codebase.
• The deployment utility would create the appropriate snowflake Users, warehouses, roles, databases, schemas, tasks, streams for a new tenant on the target snowflake account.
• The utility would establish the role hierarchies between the created roles for each tenant. Separate snowflake roles were created to grant read, write, read PII data and perform administration activities in each tenant.
• The utility would push the credentials for users created for tenants and other snowflake details to separate registration database which acted as the source of truth for these configuration data.
• The utility would copy the event coding dictionaries from S3 bucket to appropriate tenant database and make the dictionaries available to application.

(3.) SGML Data Loader

• Created a Oracle PL/SQL based utility to process manually submitted reports to PMDA in ARGUS.
• The utility would read XML files uploaded in BLOB columns in Oracle and update the appropriate tables to ensure that further reporting from can be easily done from inside application.
• Utility was integrated with OBIEE reporting tool to provide a front-end for customers to easily initiate the SGML Data Loader processing in ARGUS.

(4.) PLS Data Loader

• Created a utility in Java to load product, license and studies data into ARGUS.
• There were approximately 10,000 configuration data points across various categories that needed to be configured in each environment for the tenant.
• This reduced the PLS configuration turnaround time from a week to 30 minutes.
• The utility would generate the audit records for the PLS data inserted into the ARGUS application.

• Planned computer systems using information engineering and structured analysis.
• Developed diagrams to describe and lay out logical operational steps.
• Presented business leaders with ROI estimations to drive project planning.
• Installed system updates to address vulnerabilities and reduce security issues.
• Organized system operating procedures to strengthen controls.
• Improved systems with addition of new features and infrastructure.