How Customer’s Master Data Management Process showed a promise of 10x increase in their Productivity
About the Customer:
Client is a world leader in IT and Networking. They help companies of all sizes transform how people connect, communicate and collaborate.
Entry level pollution caused by needless data duplication due to incorrect categorizations becomes ‘Master Records,’ which are messy and consume a lot of resources to be corrected.
- Huge Duplicates
- Incorrect Data Entries
- Loss of Productivity
- Large Expenditure on Correction of Errors
- Lack of Real Time Actionable Insights
Workflow of the SmartCloud Master Data Platform
1. Customer Master Data Cleansing
- A pre-defined set of Configurable Rules
- Editable list of “Noise” characters
- Editable list of “replaceable” words
2. Customer Master Data De-Dupe
- Configurable Rules
- Classification and Clustering
- Machine Learning Derived Coefficient Values for Rules
- Master Identification
- Human Arbitration and Feedback
3. Customer Master Data Validation
- Configurable Rules
- Machine Learning Derived Coefficients Values for Rules
- Suggestions, Recommendations, and Predictions with the confidence level
- Human arbitration and Feedback
- SmartCloud teams worked closely with Cisco US and India teams for over 8 months on a very large and varied data set.
- SmartCloud team used numerous data normalization and standardization techniques (mostly rule-based and deterministic), to bring the Customer Master Data to a standard form and acceptable level of quality.
- SmartCloud team fine tuned the algorithms to make the solution generic enough.
- In the entire process various classical machine learning techniques were used to automate data preparation and metadata generation.
- Multiple algorithms were tried and the best one was chosen catering to the given data-set.
SmartCloud Data Science Platform for Customer Master Data Management was built using a Hybrid approach: Rule Based + Machine Learning + Human Arbitration.
- Client confirmed the accuracy of output to be in the 90% plus range with SmartCloud Dedupe process.
- Cisco teams confirmed that about 50% records can be auto-processed using SmartCloud Customer Master Data validation engine.
- Productivity of Client’s Data Stewards for Customer Master Data Management Process (Dedupe, Merging, Mapping, Verification, and Validation) showed a promise of 10x increase by using the SmartCloud platform.
SmartCloud Customer Master Data Management platform works efficiently with high accuracy levels for different types of Customer Master Data schemas across different geographies, customers and verticals
Advantages of the SmartCloud Master Data Platform
The SmartCloud Master Data Platform has the following advantages:
- 33% Cost Optimizations
- 14% Efficiency Improvements
- Quality Enhancements