Current State Analysis

Lume performs an analysis of current state of data. We compare data against business activities, examining correctness of data reported, and comparing data from various data sources to identify missing or incorrect validation functions that allow for incorrect data. This effort includes identifying incorrect or inconsistent formats, as well as any duplicate, orphan, or non-applicable records. This leads to documenting the current state or baseline of data quality, identifying issues and opportunities for improvement, and proposing data standardization to our client based on best practice.

Data Cleansing Plan

We analyze current state of data to inform the Data Cleaning Plan and bring data in compliance with the updated data dictionary Lume’s Data Cleansing planning activities include:

• Determining how duplicate data records will be merged, handling orphaned records, eliminating unnecessary data records, fixing format issues, and standardizing data
• Determining options of how missing required data will be handled when the data is abandoned and data owner is unviable
• Assigning severity and priority to data issues
• Methodology to obtain and reconcile data from existing sources
• Methodology to obtain missing data
• Identifying the data owner and responsible party for resolving data issues
• Impact analysis
• Defining Data cleanup environment and approach to data management
• Defining on-going plan for maintaining data quality
• Defining Metrics for Data Cleansing progress and measures of success
Developing Data Governance Plan to include:
• Communication Plan describing stakeholders and communication methods, and Roles and responsibilities
• Change control process
• Version control process
• Escalation pathway

Data Cleansing

Lume will utilize variety of data analysis, data mining, and intelligence tools to perform analysis in support of data cleansing effort, including but not limited to:

• Compiling data from multiple sources
• Maintaining, storing, and mapping data in compliance with client policies and procedures
• Updating or creating data dictionary
• Updating requirements and business rules discovered during data cleansing activities

Extract, Transform and Load (ETL)

Lume performs ETL by applying proper data modeling concepts and industry best practices to ETL during:

• Extracting the data from various sources
• Transforming data to a specified format by cleaning up, deduplicating, combining data, and otherwise getting data ready to load to destination
• Loading the data into the target database
• Testing and validation of loaded data