Tag Archives: workflow
If you are as long in the tooth as I am – you are familiar with Willy Wonka and the Chocolate Factory…one of the major plot points revolves around Charlie getting the “golden ticket” which allows him access to Willy Wonka’s factory…but there are only 5 golden tickets available.
Within the world of Master Data Management (MDM) – there is a concept of a “golden record.” This is (hopefully) not as elusive as the Wonka Golden Ticket – but equally as important. This golden record gives you access to the most pure, validated and complete picture of your individual records in your domain.
Let’s start with defining a lot of the terms from the previous paragraph:
- Golden record (according to whatis) ” is a single, well-defined version of all the data entities in an organizational ecosystem. In this context, a golden record is sometimes called the “single version of the truth,” where “truth” is understood to mean the reference to which data users can turn when they want to ensure that they have the correct version of a piece of information. The golden record encompasses all the data in every system of record (SOR) within a particular organization.
- Domain is an area of control or knowledge. In the context of MDM – it refers to the type of data you want to master. For the payer market – it is typical to start with a member or provider domain, but there can be many, many different types of domains.
One of the trickiest parts of implementing an MDM solution is creating the workflow around this golden record. You need to consider all of your data sources, which fields from which data sources tend to be more reliable and what are the criteria for allowing the field from one system to populate an MDM field over another. In other words, if you have an enrollment system that captures the member’s name and a claims system that also captures the member’s name – which of these two systems tend to have the most correct member’s name? Is there another source system that is particularly reliable for capturing address – but the member name tends to be off?
One of the main considerations in the creation and maintenance of the golden record is matching and merging records. If there are two records that are pretty similar, what is the process for inclusion in the golden record? – for instance consider the two records below:
|Last Name||First Name||Member #||Phone Number||Street Address||City, State|
|Wayland||Jennifer||201215||7065842||123 Maine Street||Camden, Maine|
|Wayland||Jenn||201211||2078675309||123 Main Street||Camden, Maine|
The last names are the same as are the City, State fields. All of the other fields are different – in the world of a golden record, they don’t create an automatic match/merge. For the sake of this example – let’s say that we know that the first record comes from a source that has great reliability for names and addresses, while the second record comes from a source that is known for highly accurate member numbers and phone numbers. With good MDM solutions there should be a toolset that allows you to automate the merge functionality as much as possible. For this example, you could set up the workflow to obtain the records from the two sources, set up the criteria for merging/matching (take the name from the first record, the member number and phone number from the second record, and the addresses from the first record. The following could be the golden record for this member:
|Last Name||First Name||Member #||Phone Number||Street Address||City, State|
|Wayland||Jennifer||201211||2078675309||123 Maine Street||Camden, Maine|
Where matching and merging get interesting is when the source fields are not clear “winners.” There will be situations where manual intervention is necessary to determine which record should take precedence. For this – a workflow manager toolkit is very helpful. It will assign records to a data steward who can then make the judgment based on their specific experience and knowledge of the data set which field from which record should take precedence. It can also have approval mechanisms before a record is finally and truly merged resulting in a modification to the existing golden record for a specific member.
As a result of the complexity of implementing a Master Data Management solution – it will help to start with picturing your golden record. Can you answer the following questions?
- What information needs to be captured in your golden record?
- Related to this – is there any information that is not necessarily specific to the domain but may be interesting when attaching relationships to a record (attaching a provider to a specific member)
- What are all the sources of data for the record?
- Are all the sources currently integrated? How easily can new records or updated records be shared?
- Which source is the best source for which fields?
- What is the threshold your company can tolerate for automatic merges?
What approval process needs to be in place before a merge takes place? Who needs to look at the record/recommendation before the merge is complete?
By Nancy Atkinson, Senior Analyst, Aite Group
Karen Hsu of Informatica organized a TweetJam (#INFAtj) recently on business-to-business (B2B) payments, SEPA, and integration. In conversation with Chris Skinner of Balatro Ltd., I stayed (mostly) within the 140-character message limitations of Twitter while the hour flew by. (more…)
What are the opportunities for integrated solutions to address the health industry gaps discussed in prior blog postings? One of the exciting possibilities is the use of IT to make the overall health care system more efficient through a new class of intermediary – Health Information Exchanges (HIE) – that serve as integrators of the highly fragmented information silos that exist in the industry. HIE’s are, in essence, Integration Competency Centers that promise to improve health care delivery by improving efficiency and applying Lean Integration principles. (more…)
In my last post I introduced the concept of Fedegration as the ability to leverage and manage data and content across applications for new purposes and still leave it in place for use with the existing application.
I will now discuss how our team has built and integrated what I think is the best set of technologies for a Fedegration Platform.
Let’s hit the requirements first. I believe an ideal platform should: (more…)
The recent blog post from Forrester Research analysts Clay Richardson and Rob Karel posed an excellent question around synergies between Business Process Management (BPM) and Master Data Management (MDM). Siperian customers have been at the forefront of driving these synergistic requirements that have made the Siperian MDM Hub uniquely suited for bringing together BPM tools such as Lombardi with MDM to enable data governance.
Together with the excellent points made by Clay and Rob, one of the trends we have seen is that most of the integration of MDM and BPM has been focused on the inbound – the creation of master data. But in the outbound – synchronizing master data (from the MDM hub) with downstream systems – BPM has been less leveraged or involved.
Inbound: While an MDM hub automates the merging of a large volume of duplicates, exceptions need to be handled by the data steward in "collaboration" with the data owner/ business user. Additionally, business users are starting to interact with Hub data directly as a “system of entry”, through interfaces like Siperian’s Business Data Director. Both of these inbound scenarios require a “chain of approval” and potentially sophisticated rules for process flow. A BPM tool such as Lombardi integrated with Siperian and leveraging the Hub’s ability to store states (transitional states of records prior to being finally committed to the hub). See a previous blog post The Art of MDM Workflow for more information.
Outbound: Downstream systems that accept the data from the MDM Hub can automatically receive updates through data synchronization. Once the master data is created or updated in the MDM Hub, it can place records onto message queues for EAI style distribution to systems such as CRM/ ERP, allowing those applications to have up-to-date accurate master data. Alternatively, periodic batch exports and updates using ETL can also accomplish this albeit in a non-real time manner. Typically, since there is no human interaction or complex process flow, BPM does not enter into the equation for such processing.
In summary, so far we have seen great use cases for using BPM inbound, not so much on the outbound side. What have been your experiences and do you have situations where you have applied BPM on the outbound?
The art of war is of vital importance to the State. It is a matter of life and death, a road either to safety or to ruin. Hence it is a subject of inquiry which can on no account be neglected
– Sun Tzu, the Art of War
As organizations begin evaluating and leveraging MDM beyond their analytical needs and start to view operational MDM as a cornerstone of their overall strategy, the question of governance and workflow becomes a vitally important topic, which cannot be neglected. For many, workflow and business process management (BPM) is a well ploughed road, with established vendors such as Lombardi (a Siperian partner) providing robust offerings. So the marketplace would have you believe in a simplistic view
Workflow Tool + MDM Hub = Victory and Mission Accomplished
but the “Art” of MDM workflow actually lies within the subtle, yet flexible nature of the MDM platform itself, the ability to handle state. Not the State as referred to in Sun Tzu’s quote but the state of the data or record as it progresses through a business process towards its ultimate destination. A significant concern for organizations looking to enable their MDM Hub as a gateway for system of entry of all master data is the ability to enforce the right level of governance through checks and balances as additions and updates to data are made by business users. A workflow tool functioning independently is designed to easily track and route the operations and approval processes to the correct defined individuals or groups. However, a key requirement of the MDM Hub, in concert with such “in-transition” workflow activities, is to be able to manage the changing state of the data.
For example, when new data is created, the MDM Hub should be able to track its state as “pending approval” or any number of additional states before its final destination – becoming the golden master data record. This built-in capability should permit complete control over the availability of the data, framing it as not yet blessed for enterprise-wide public consumption, but continuing to allow the Hub to determine and resolve conflicts that might occur. In our example, if a second addition is attempted for the same record while the previous entry is currently in the approval process, the MDM Hub’s matching capabilities would highlight such a conflict, thereby preventing duplicate effort or accidental overwriting of data. Therefore, in order to succeed, the workflow tool must leverage the unique data state management capabilities of a MDM Hub. Only then can we claim victory and mission accomplished.