BLOG

Bringing the promise of ML to your MDM : Part II

Thomas Varghese
Thomas Varghese
Analytics/Product Consultant

‘Augmented data management’ is a key trend where AI/ML is transforming how enterprises manage their data.

In the last article, we looked at some of the key pain points that exist as IT and business leaders constantly grapple with the increasing influx of data sources, without systems to keep up.

Let us look at what makes AI Data Cleanser address these pain points.

So what is AI Data Cleanser?

AI Data Cleanser is a suite of AI/ML based data management tools that aims to deliver reliable data to your business.
The image below illustrates the breadth of issues that typically exist, and the specific entities/use cases the solution addresses.

Let’s look at why each of these use cases is crucial to tackle from a foundational perspective.

  • Data validation – Your business units constantly refer to master data as a ‘source of truth’; it could range from critical data such as customer shipping information, a lead’s contact email or employee phone number. Maintaining, updating and constantly checking data validity to ensure business sees the right information is a key determinant of the quality of downstream decisions made.
  • Data cleansing – Enterprises rarely have their master data in one source – standardized, cleansed and ready to go. The reality is that each data entity is sourced from traditional platforms (CRMs, ERPs), flat files, external sources, etc. and this often leads to redundant and duplicate information. AI Data Cleanser leverages powerful machine learning models to identify similar entities, group them and assign a representative ‘golden record’ that helps the business identify and tie all relevant information to that unique record.
  • Data enrichment – Every firm is well on its way to using internal data for decision making; however, the wealth of information present outside your firewall could help provide key insights on multiple fronts. This is where firms are keen to compete and gain a competitive advantage. As an example –
    • What if you could tie each of your customers to their parent firms, and actually identify white space opportunities to grow your business?
    • What if you could validate and enrich your product attributes, while also analyzing relative assortment and competitive pricing trends on e-commerce sites?
  • Hierarchy management – Hierarchies are a tough nut to crack, as they often combine the problems of the above use cases, and add complexities of their own. However, a robust hierarchy mapping of your customers, contacts, products and materials can be invaluable in gaining a 360 view of your business and providing opportunities to grow revenue while controlling cost.
    • An interesting application of hierarchy management is product category standardization (in this case, to the GS1 standard), which we have implemented for a few large retailers in the EU. This migration helped our customers streamline product lines, optimize their supply chain and rationalize their supplier portfolio.
  • Data anomaly analysis – “What is the state of your data quality”, “What are your data quality challenges”?; These questions could either prompt a blank silence or lengthy answers without a clear direction. The reality is that data quality metrics in any firm is complex due to the multiple issues we have established. However, solving any of the use cases we have seen above without providing business users and IT teams with custom reporting and insights into their data health is only solving part of the problem. AI Data Cleanser leverages ML driven anomaly detection tools to test variations in data, and also allows custom business rules to be defined; thereby leveraging clear data quality standards to be tested, in order to measure and improve data quality.

How does AI Data Cleanser work for you?

  • Data monitoring – Here is where we close the loop on the feature set. AI Data Cleanser is built to integrate into your environment and run projects on an ongoing basis. As user feedback is provided, model accuracies increase, thereby improving automation and data quality KPIs. AI Data Cleanser is configured based on use case(s), and comes with a managed services team that works to scope out client specific requests and enhancements that are to be built as part of the project.
    • The advantage with AI Data Cleanser is its modular plug-and-play model, where different use cases use single or multiple components of the solution.
    • For example, a use case to cleanse and master customer data from Salesforce would have a workflow quite different from a use case to validate customer addresses and build customer hierarchies.
    • In other words, the solution is priced and deployed as per customer needs and integrated with the systems/processes they use currently.

AI Data Cleanser connects to a range of input systems, ingests data through a “data discovery” layer, where data is unified and standardized, and then processes data based on the configuration defined.

The solution is cloud compatible as well as on-prem friendly and presents multiple options for integration.

AI Data Cleanser has seen a number of successful implementations with varying scale, right from a simple customer validation and de-duplication exercise all the way to replacing an enterprise MDM platform for address validation.

In the last part of this series, we will explore some of the different flavors of the implementations done through AI Data Cleanser.

Learn more about AI Data Cleanser, and reach out to our team for a free demo – www.tredence.com/ai-data-cleanser/

0 Comments

Submit a Comment

Your email address will not be published. Required fields are marked *