When I was reading the most recent Gartner Magic Quadrant for Metadata Management Solutions, I noticed that Silwood Technology have a relationship, either commercial or technical or both, with 9 out of the 12 vendors in the Leader’s quadrant. We also work with vendors in other quadrants.
Our software product, Safyr® provides them with a quick and easy way to provision their metadata management solutions with metadata from large, complex ERP and CRM application packages. I do not know how the other 3 achieve this.
To illustrate the value that those vendors and their customers receive it is worth looking at the experience of one of our first customers with Collibra who resell Safyr®.
This is ATB Financial in Canada. Using Safyr’s interface to Collibra they were able to import thousands of tables of SAP metadata into Collibra quickly and easily. Wellington Holbrook, ATB Chief Transformation Officer said, “We took what would have been months of work (and possibly a barrier to progress) and completed the activity within hours and nominal resource investment. We’re pleased to have been able to achieve a high degree of efficiency with this collaborative effort.”
In fact one of the those in the Leaders quadrant contacted us in 2018 and asked us to build an interface from Safyr to their product as they could not easily bring metadata from SAP and other ERP/CRM packages into their solution. Obviously, we were delighted to help!
This is a common problem which can delay or even derail an enterprise metadata management project. Without the ability to quickly and effectively incorporate relevant parts of the the data model from your enterprise ERP or CRM packages into your metadata management solution, it will not be fulfilling its potential to transform the way you use data.
Essentially most of these packages share three characteristics which mean that traditional methods for scanning their metadata and provisioning it into metadata management products are ineffective to say the least.
Most ERP and CRM packages are based on a relational database platform so you can be forgiven for thinking that it would be fairly straightforward to use a database scanner to pull out all the table and attribute names, description and table relationships.
In principle this should work, however, there are two features of ERP and CRM packages which make this an unworkable solution.
The first of these is that there is no meaningful metadata, by which I mean ‘business names’ and descriptions for tables and attributes and no table relationships defined in the database system catalog.
The result is, that even if a data catalog scanner reads the schema of the application database, the results are of little value. You would be faced with trying to decipher what the physical names for tables and attributes mean as well as how tables are related without any information about primary and foreign key constraints.
The second is that ERP and CRM packages have very large data and complex models. This adds another layer difficulty to locating what is important to you. For example, SAP has over 90,000 tables and 1 million attributes, and a typical Oracle eBusiness Suite implementation has over 20,000 tables before customisation.
Whilst we appreciate that metadata management solutions can cope with that level of scale we do not believe it is practical to provision them with that quantity of metadata especially if it is not organised so that business users can find what they need quickly and easily. For example, it is quite possible for many thousands of tables to be unused which means that they may be unnecessary for the metadata management project.
In almost all packages, the useful metadata (business descriptions, relationships and more) is held in specific groups of data dictionary tables. It would of course be possible to query those tables and store the results in say, a spreadsheet. However even doing that leads you to the next problem which is how to organise and search that metadata to find the subsets which are relevant.
As it is unlikely that scanning and importing so many tables into a data catalog’s metadata repository is necessary, valuable or even viable it is important to be able to subset the metadata into relevant and useful business artefacts for your metadata management project. Examples are subjects such as “bill of materials”, “sales order”, “accounts payable”, “vendor master” etc.
Without a specialist product which organises the metadata so that you can search and subset it quickly and easily you are limited to using some vendor tools which are not designed for use by data analysts and scientists.
Alternatively, you can employ consultants to try to do the work for you, or try to use templates or whatever documentation may be available – taking into account of course that your system may have been customised during implementation.
Whatever method you use to create the subsets, once they are complete you will want to provision them into your metadata management environment.
If it is not in machine readable form then the only options you have are to hand key it or perhaps copy and paste the information into the data catalog or into an intermediate file or other staging area. This is time a consuming, costly and potentially inaccurate method.
Imagine for example trying to rekey the information from a single SAP table, MARA (General Material Data) into a data catalog. It has over 240 attributes (columns) each with a business description and is related to over 1500 other tables in the SAP system. We normally estimate that it takes about 1 day to rekey the metadata from 5 tables accurately into another system.
Clearly you will need metadata from more than a few tables in your data catalog so this is not really a viable option.
The quickest, most effective and accurate method is to import it using whatever mechanisms are available. Depending on which metadata management you are implementing, this may be via an API or importing files in various formats.
However, this is not necessarily a simple task, especially when the metadata is complex as you will also need to understand the data structures of the metadata management in order to know how to create the source to target mappings. If you have large quantities of metadata, there may be issues with speed of processing.
And, if you have not been able to access and analyse the business definitions and relationships in the application metadata then only loading physical table and column names into the data catalog is of dubious value anyway.
Safyr® is a single product which is designed to help you to overcome those hurdles in provisioning metadata from complex ERP and CRM applications to your enterprise metadata management solutions. You can use Safyr with SAP, S/4HANA, SAP BW, Salesforce, Oracle eBusiness Suite, Siebel, PeopleSoft, JD Edwards, Microsoft Dynamics AX 2012 and Microsoft Dynamics CRM.
Safyr saves you time and money when compared with the traditional methods of achieving this.
Safyr is a specialist self-service metadata discovery product for ERP and CRM packages. It ensures that your enterprise metadata management project delivers a comprehensive and inclusive solution that is populated with rich metadata from these complex systems.
Safyr accesses and extracts the technical and business metadata from your ERP and CRM systems as implemented in your organisation, including customisations. This means that you are always working with accurate metadata.
You can then use Safyr to search, analyse, subset and curate that metadata based on your business needs. Safyr then delivers the selected metadata into the enterprise metadata management product.
You must be registered to receive support
Please login or Register Here »
Note: The token is a unique string for each Safyr license and can be found on your product delivery instructionsAlready registered? Login here »
Enter your username
to receive your password