Master data management: Creating golden records for business
Estimated reading time: 4 minutes
Nick Lim, TIBCO’s General Manager in APJ, confesses there has been a lot of significant change for the organisation in terms of strategy, technology, and innovation.
Table of contents
TIBCO CEO Dan Streetman’s clear vision since 2019 to solve the world’s most complex data problems is based upon the company’s very strong legacy solution in integration.
Nick explained, “We provide integration solutions (or middleware) to a lot of enterprises, and almost every bank is using our solution, along with manufacturers, and even some telcos.”
Besides integration solutions under its Connect product pillar, there are also core product pillars of Unify and Predict, which Nick wanted to highlight can help companies ensure consumers’ personal data that is disclosed to create personalised experiences, is not abused.
Connect, Unify, Predict
Integration solutions come under TIBCO’s core product pillar called Connect, which Nick describes includes all integration via APIs (Application Programming Interface) between different systems, for example application to database, or mainframes to the cloud, and so on.
After that comes the data governance portion (under Unify) to make sure the data is accurate. Here, data is transformed into master data which can be fed into visualisation tools or data science tools (under Predict) to create data-driven predictions and decisions.
He said, “To us the Connect portfolio of solutions we have is about integrating or linking infrastructures together, in the most efficient way.”
Of course, the main objective to do so, would be to facilitate exchanging of data.
With three (or more) different BUs possibly collecting the same data about a single customer, how does an organisation consolidate all of it and ensure that what is consolidated is cleansed, governed and is the most accurate data, or the golden record?
After that comes the data governance portion (under Unify) to make sure the data is accurate. Here, data is transformed into master data which can be fed into visualisation tools or data science tools (under Predict) to create data-driven predictions and decisions.
The solutions under these product portfolios all aim to solve challenges that the current data landscape faces today – too much data, too many siloes of data, and the complexity of consolidating them all.
Challenges
Organisations are having a big challenge trying to make sense of all these data, Nick said. On top of that different applications collect different data. “Insfrastructures are in siloes, and business units (BUs) are in siloes.”
With three (or more) different BUs possibly collecting the same data about a single customer, how does an organisation consolidate all of it and ensure that what is consolidated is cleansed, governed and is the most accurate data, or the golden record?
That’s where the solutions Tibco has under the Unify portfolio, comes in.
Nick also cautioned that it is not only data and the infrastructure that need to be integrated or connected but people and processes also have to be. which is quite a challenge.
Best practices
Nick said that customers need to understand the context of their data, as well as understand what they are collecting.
“I think the business analyst or the business owner need to work very closely with the IT to make sure that they are collecting the data that is necessary.
A second practice Nick shared is to only collect the data that you need.
Ultimately, a golden record should be created, for example a single version of truth by an organisation about a single customer, instead of three different versions by three different business units.
“Personally, I think this will solve a lot of security problems as well.”
The rationale behind this is that when you do not over-gather data, data becomes easier to manage.
Nick also encouraged organisations to have an established process between the business user and the application or IT department. “They need to communicate frequently to ensure the data can be interpreted as accurately as possible.”
Ultimately, a golden record should be created, for example a single version of truth by an organisation about a single customer, instead of three different versions by three different business units.
“If you can do that, and you do it well you have mitigated risk of security issues.”
Besides that, golden records are located in a single place and centrally managed.
This central control ensures only the right people have access to the data which they are supposed to have access to, via role-based access control features.