In what circumstances would need to use ODI as a complement to a deployment of EDQ?

Question

Im about to do a presentation about Oracle Enterprise Data Quality and Oracle Data Integrator.
In my understanding if I use ODI  to do ETL/EL-T tasks, ODI needs EDQ jobs to perform advance Data Quality procesing. In that escenario is perfectly understandable the implementation of both ODI and EDQ solutions.
If I does not have any solution to do ETL/EL-T task, and my objetive is mainly do data cleansing, (take a set of data, cleaning it, and write the results on a database) why I would need ODI? There is a feature that I could need from it, that I do not have in EDQ for this purpose?
This doubt came to me, becouse I found a lot of references that makes me think that this two solutions are codependents.
Thanks for your reply!
regards.

Answer 1

Hi,
This might help. Note that ODI and EDQ are normally sold together for Data Warehousing scenarios, though arguably data profiling, at least, is important in nearly all new projects that use ETL/E-LT technology.
Some customers, especially MDM customers, do use EDQ for their ETL requirements, as these tend to be simpler than data warehousing scenarios where load plans and data aggregations are important. ODI also has more comprehensive connectivity than EDQ, for example for Mainframe systems.
We are also seeing a trend for ETL/EL-T to be done strategically in such a way that it rarely needs revisiting for new projects; for example when a new system is onboarded, the key data is replicated or ETL'd into a standard holding area which is then used for any projects which require data from many source systems. This means that EDQ very often takes data output from ETL/EL-T without necessarily being embedded in the ETL or data replication flows.
Very often both ETL and DQ are required
It’s not the size or complexity of the transformation that is different, it’s the nature of the transformations – syntax or meaning?
ODI will do all sorts of denormalisation/normalisation for populating star schemas that would be complex and slow in EDQ
EDQ will do all sorts of things that improve the business value of the data that ODI would never do
Data volumes are unlikely to be a guide as to which toolset would be used for a particular task
Capability 
Data Quality tools (e.g. OEDQ)
ETL tools (e.g. ODI)
Extraction of data from common types of modern database / common file types
Yes
Yes
Native extraction of data from antiquated systems using specific APIs, and data from complex file types, e.g. XML, EBCDIC, EDI, IDoc
No
Yes
Complex extraction rules
No
Yes
Syntactic low-level transformations
Yes
Yes
Data profiling
Yes
Basic or none
Semantic understanding of data
Yes
No
Complex transformations using semantic rules
Yes
No
Complex transformations using non-semantic rules (aggregation, analytical, etc.)
Possible, but ETL usually preferred
Yes
Complex matching and merging
Yes
Limited (ETL-type matching/merging) such as exact ID match
Processing as native SQL on database for optimal performance wherever possible
Some
Yes
Advanced performance parallelisation / distribution / control
Rare
Yes
Multiple source/target mapping
Some
Yes
Writing of data to target systems / staging areas / files
Yes
Yes
Advanced load capability, e.g. single/packet/bulk, load plans etc.
Rare
Yes
Typical users
Business Users, Consultants, or IT Professionals
IT Professionals
Regards,
Mike