K. Nisha Sreedharan

Projects/Responsibilities

• Created custom Data Governance module using PySpark, Airflow, Hive and Pandas and reduced manual efforts for each business unit by up to 30% and increased data quality controls for 10 processes
• Automated several data pipelines for supply chain and spend business units and thereby reduced ~80 hours per week of manual effort
• Worked on setting up data quality rules for different business use cases on Amex Enterprise Data Governance Platform
• Managed PMO activities for Spend ChatBot
• Built dashboards on Tableau

Projects

Ocrolus is a fintech company that helps customers make faster and more accurate lending decisions .

• Set up pipelines to bring data from tools like Salesforce , Airtable, Ordway into Redshift and Snowflake. Using Python and AWS
• Setting up alert systems through Redash
• Snapshot project to send usage reports to clients . Developed using Python and AWS and helped customer satisfaction for 60% of accounts.
• Code for role based access in Redshift . Developed using Python and AWS
• Built dashboards on Redash for Volume Forecasting

Projects

Goals101 is a product based 'Transaction Behavioral Intelligence' company that helps to bridge the gap between brands and banks

• Data Pipeline - A robust pipeline built to automatically ingest data daily and perform quick data quality checks. Project built using Airflow and Pyspark and Grafana
• Audit Framework Project - To automate data validity checks on all KPIs for data shown in front end vs data in database. Project built using Elastic Search, Pyspark
• Brand Planner - To detect seasonality trends of any brand. This module can be used to push offers for a brand in a particular month based on past performance of brand. Built using Spark, Scala and Elastic Search and Airflow
• Spotlight - This tool is used for brand specific insights as well insights regarding peer brands with similar spend patterns to analyze yearly /monthly spend patters and customers etc using Spark
• Normalization Framework - Framework developed using PySpark to convert raw data received from banks by choosing various transformation functions
• Location Parsing - To iterate raw narrations received from bank and segregate data based on certain regex expressions. Project developed using Spark
• Anomaly Detection Module - To detect transactions with unusual spend patterns analyzed over a period of time. Project developed using statistical models in PySpark

Projects

A product based company that specialized in building a generic data lake.

• Data blaze data lake - Developed in Spark , Hive , Cassandra
• Insurance fraud detection - Medical Providers - Developed using Spark, Cassandra,Hive. Increased fraud detection in the claims with 75% accuracy.

Projects

• Data Ingestion Framework - Data pipeline using Spark, Sqoop,Oozie, Hive
• Data Quality Framework - Developed using PySpark,Java and Hive
• Insurance Fraud Detection - Using Spark Streaming to calculated weighted averaged based on conditions
• Cluster set up ( VM and HDP Ambari)