We at Kitepipe have been building solutions with the Boomi Master Data Hub for quite a while. In fact we are leaders in the use of the Boomi Hub technology in solving data mastering problems for Customers in the Employee, Product, Customer, Vendor and Contact domains.
Like all Boomi modules, the Hub is a flexible toolkit that can be applied effectively to many situations, but requires experience to deliver high function designs, I recently queried the Kitepipe Technical Architects for Boomi Hub design tips - here are some of our best learnings:
One: Don't Include All Useful Data In The Mode
Don't include all useful data in the model - just the attributes that must be synchronized across platforms - with APIs and integration, you can go get additional data from “source of truth” application as needed. This speeds up implementation and reduces time to value.
Two: Put Reasonable Maximum Character Sizes On All Your Text Fields
Put reasonable maximum character sizes on all your text fields. You can run out of space in your model if you let every text field use up the default large number of characters.
Three: Build & Use Launcher Processes
Build and use 'launcher’ processes to sequence the feeder processes properly. Relational links in the model will require that the spoke process run in a certain sequence. And, build them using Process Route shape so that you don't have to re-deploy the whole launcher when one spoke needs an update.
Four: Start Small & Iterate
Start small and iterate, learn to leverage the MDH configuration options to drive functionality (rather than fighting it).
Five: Use A Connector License for The MDH API
Plan to use a connector license for the MDH API - it is separate from the MDH connector, but exposes many powerful features that greatly enhance success and usability.
Six: Make A Reference Domain That Stores Name:Value Pairs
Make a “reference” domain that stores name:value pairs for categories and picklists that are used to match data values across applications. Just takes a small number of Golden Records, and is much more maintainable than keeping cross reference tables in the spoke processes.
Seven: Don't Underestimate The Value of Clean Data
The primary barrier to success is the data quality in the applications. Don’t underestimate the effort to initially clean the data, and the ongoing maintenance required. “Data quality and data stewardship is a journey, not a destination” said a wise Kitepiper.
Thats seven - we have learned many, many more tips and techniques. Why not contact me and I’ll arrange a MDH model review with a Kitepipe Technical Architect, so we can share some of our MDH design secrets.
