I was recently approached to give an interview regarding Enterprise Mashups. The main thrust of my answers shows my thinking that capitalizing on Enterprise Mashups and SOA requires IT to undergo a fundamental shift in mindset. IT must stop solving problems and start enabling end users to solve the problems themselves.
Here are the questions along with my responses.
Could you give me a quick background for yourself? What’s your current job?
I currently work for Aera Energy, LLC. Aera is a large, privately held oil company located in California. I work as a Data Warehouse Development Lead and Web Technology consultant. I also do consulting work for other firms.
Disclaimer: The statements and opinions in this blog post are my own and do reflect the opinions or positions of Aera Energy, LLC or any other firm for which I do consulting work.
Have you been involved with mashups long?
I’ve been involved with web mashups for about 6 months. I have done other forms of web development (database backed web sites, document management systems, web email systems, etc) for 11 years.
What are the key concerns that make enterprise mashups unique?
Enterprises have struggled with disparate databases and disconnected applications ever since IT became a fundamental part of doing business. There have been many different ways to solve the problem like EAI, EII, Data Warehouses, CORBA, DCOM, web services, SOA, etc.
Enterprise Mashups are basically a new way of solving the old integration problem.
They are uniquely positioned in their ability to quickly solve the integration problem because they put the problem-solving power into the hands of the end user. The technology isn’t really the big enabler — the big enabler here is the fact that your average end user is becoming much more technology savvy.
This trend will only accelerate.
The employees of tomorrow will have grown up with the web and internet as a part of their daily lives. The next generation will be savvy enough to make their own applications. Many employees of the current generation are already doing it. IT must become more of an enabler than a solution provider. (I wrote about this in another blog post regarding programmers and enterprise mashups.)
Do you think that the new tools such as Yahoo Pipes and QEDwiki are going to be useful for end-users in an enterprise environment, or are they just going to make the IT guys’ life easier?
You can divide users of tools like Yahoo! Pipes and QEDWiki into three groups:
- end users
- power users
- developers
Yahoo! Pipes and QEDWiki are great tools for developers and power users. They are also simple enough that many people who currently fall into the group of end users may jump up to the level of power user. So, yes I think they will be useful for end users.
In terms of making the lives of IT guys easier, I actually think they’ll make life harder for IT developers. IT developers are going to have to shift their mindset from solution provider to solution enabler. They will have to design their systems as services to be consumed by other people. This is similar to the problem faced by software library developers in the past — it’s much harder to make a reusable system than it is to just make a standalone solution.
What sort of IT capacity should companies develop to ensure that can best utilize mash-ups in the near future?
IT needs to develop the solution-enabler mindset. Having that mindset is much more important than having a particular piece of hardware or software.
SOA goes hand-in-hand with Enterprise Mashups, so efforts to develop an SOA will be key to capitalizing on Enterprise Mashups.
I think the best way to get started is to look at the home grown solutions that currently exist in your enterprise. Many times these solutions will be complex Excel spreadsheets, advanced Access applications, or internal web sites. Talk to the people who maintain those applications. Figure out what problems they are solving and what technology and data hurdles they have to overcome. Those hurdles point to places where IT can become more of an enabler than a problem solver.
While you can obviously design an overarching architecture I think it would be a mistake to start that way. Your first attempt at SOA and Enterprise Mashups should be small. You’ll probably get it wrong the first time so the important thing is to get started, learn from your mistakes, and then make something great.
enterprise2.0,soa,enterprisemashups,enterprise2.0,eai,eii