One of Alfresco's strengths is the data modeling. Document domain experts can model type and hierarchical associations and metadata constraints in a declarative way (such as an XML definition).
Moreover, Alfresco added a graphical user interface model manager in the last product version, so users that don't know XML formalisms are able to build models.
Alfresco supports category concepts, so users can categorize all documents and add expressiveness to data and conduct more powerful searches.
Alfresco services can be extended in a simple and well-designed way, such as Maven and pre-designed Alfresco module package Java projects. Learning paths are well documented and developers can add your new ECM services at the top of the platform.
The Alfresco community source code is public, so developers can study it or extend it in the desired way.
We would like to see improvement in the numbers of documents that can be managed by the repository. We would like them to introduce NoSQL DBMS for modeling metadata and document types.
We have used the Alfresco ECM platform since 2009.
For large repository sets, the architecture must be well sized.
The product uses state-of-the-art J2EE frameworks and libraries.
I rate technical support a 3.5/5.
Involve an Alfresco partner for the project. The training is basic. Buy Alfresco certifications and boot camp trainings.