Participants:
Rasmy, Sherif
Yelton, Ken
Sears, Andy
Winters, Jonathan
Carlson, Doug

Key Points:

Speaker Name: Sherif

Improve the integration of systems together in terms of handling low level data access, asynchronous and synchronous messages, help aggregate data from different sources and present them in a single model.

ESV can be used to come up with an aggregate of information from different systems.

Data-warehousing

Analytics… collect as much information about the entity/subject upon which you would like to investigate. It is important to collect a lot of information.

Real-time synchronization.

Mule / Camel / Spring – standard design patterns, proprietary vs. open source.. Mule has documentation, literature, maturity.

CRM  (Telisma?) – customer relationship management of students, constantly / real time extracting

If you do not own the system, you have to deal with security issues as far as other groups opening systems for a connection.

What are services that we are looking for…  get 360 view of customer.

 

Andy Sears:

Ok from boss to explore…. “what is one place that can have all data, anywhere at the University that can be exposed through one area…. aka an NYU API”

Sherif Rasmy:

Different schools can have different student ID’s. Internally schools have different ID’s. FAFSA ID, etc.

Data-warehouses, collecting all information. Not scalable.

Martin Fowler … book about integration patterns.

Ken Yelton:

Why did we choose Talisma instead of SalesForce. Not responseive enough.

Sherif Rasmy:

Every department may have a set of web services or applications.
Piece of software to handle connectivity, security, transaction, load balancing, scaling. ESB’s allow less and less code for data exchange.

Jonathan Winters:

I think this is indicative of the nature of how our University works. Each IT decision seems to be made on a small scale, within departments or at the school level. WHen developing service offerings  and applications perhaps they can be done with the University-wide scope in mind to begin.

Ken Yelton:

My team is working on personal ID data across all systems using Oracle.

Andy Sears:

Grouper could be an option. Creating groups across the enterprise. Does anyone use it on the NYU campus.

Sherif Rasmy:

Concept or object of a person. How do we aggreage… from peoplesoft…

“Golden Source” hard to rely on because: Hard to collect all information, different systems store different types of products,  firms started to do it, many systems that incrementally have to be put into system, application architect can not depend on golden source because it is not complete and when you want to use it, it is going to be late.

Ken Yelton:

When we make an update to data on one system, there is a system that defines where it is going to go.

 

Sherif Rasmy:

You could have a couple of ESB’s talking to each other, better than having none at all