Sanjay, for the readers in our audience who are not familiar with Izenda, can you tell us Izenda areas of focus?
Sanjay Bhatia: Izenda is the market leader in self-service reporting, real-time dashboards and data virtualization. We’ve been around for about 10 years. For the past few years, we’ve integrated into more than 1,000 applications. Essentially we let users create and customize their own reports. We found that typical users need the ability to make numerous simple customizations, and that’s currently something people think of as an IT function. Our vision is that reporting is something that belongs in the hands of the user community. IT is there to secure it, but once they do that, they need to hand over the reins so users can create all the reports they need. Sometimes they need things very quickly so the self-service model is really the future.
I certainly agree that allowing the business users the opportunity to make simple reporting changes and reducing the burden on IT is a win for everyone. Can you define what you mean by data virtualization and explain why it’s becoming more important from a business intelligence perspective?
Sanjay Bhatia: Data virtualization is basically connecting many data sources simultaneously and making it seem as if you’ve copied everything through an ETL process. But in reality, the data is stored in the source environment or application. This is important because data volumes are getting huge, and the hard drives, storage and processors processing that data are getting more powerful, but the networks and the bandwidth to move all that data are not increasing nearly as fast. So you get into situations very quickly where it’s really not possible to move data overnight to a data warehouse. Aside from that, it tends to be an 18-month initiative, and by the time you’re finished with it, a lot of enterprises find that the original requirements have changed and the product no longer suits what you wanted to do.
Also, from a data quality perspective, data virtualization gives you access to real-time data, not a snapshot. This means that users can look at a dashboard, drill into the data, make a change – whether that’s a data problem or an actual operational problem – and see the result before it gets on anyone’s radar versus a traditional data cleansing process that happens once a quarter.
Let’s focus a little on data quality. How does data virtualization affect traditional data quality practices?
Sanjay Bhatia: It really incorporates the Agile methodology – the idea that we deploy a small amount of data to the user community, they’re able to look at it. When there are problems, they’re given the opportunity to make rapid changes. It serves the purpose of no longer allowing dirty data in the operational side. In the application, users have the expectation and the tools now to look for data problems and correct them. This does not replace the traditional data cleansing or audit process, but it’s probably one or two orders of magnitude simpler and less time-consuming. We don’t make a clean copy of the data – you just basically correct operational data. There are no “clean” or “dirty” copies anymore. You just have nothing but good data.
So it’s really focused on the business processes where companies are operationally connecting with their customers. Is that a major point of emphasis for you?
Sanjay Bhatia: Yes, especially from the compliance side with Sarbanes-Oxley and Dodd-Frank Act on the horizon. There’s a real initiative out there from the government and from a competitive global perspective to make sure that data is compliant and that if something is a problem, it doesn’t wait until the end of the quarter to be fixed. This also is important from the perspective of the news cycle – and this is something that is becoming faster all the time – so that a problem doesn’t escalate to the point where there may be media exposure. You actually can get the problem taken care of very quickly before it snowballs.
We hear a lot today about “agile” and there are many different definitions. What does making data virtualization “agile” mean to you?
Sanjay Bhatia: It means you can literally deploy a virtualized data warehouse over the weekend, perhaps with a small set of functionality and requirements, get feedback and gradually increase the data that’s available to the user community instead of using the waterfall model of defining everything you want to do. That rarely works because people often don’t know everything they want until the second they need it. Or if you’re doing a typical 18-month business intelligence (BI) project, by the time it gets deployed, the original requirements are no longer relevant to business – a country may have gone bankrupt or one of your competitors may have been acquired. Product cycles are shorter than your typical BI time span to deploy something. So agile BI to us is really about taking the operational databases, plugging in Izenda Reports, and delivering that very quickly – in days rather than weeks or months. You may not give users everything, but you give them something and based on their feedback, you can prioritize what the next step should be – what data sources you’re going to reveal, or whether aliasing or data aggregation needs to be done. That way you deliver value right away, and you don’t focus your energies on the wrong tasks.
How are you seeing the reporting function changing in organizations today?
Sanjay Bhatia: I think it’s part of a bigger trend. Twenty years ago, the Internet was something that UNIX administrators with command line prompts were using, and that’s the only way you could access it. Today you literally see children with phones taking pictures and videos and uploading them to their blogs for their families to see what’s happening in school. People feel very empowered, especially with tablets. It’s something that’s very natural. I think BI and reporting are definitely going in that direction.
We really don’t see IT as the key holder to the data. We see that, yes, there is some security and compliance that they have to deal with, and they have to make sure the reporting capability – like our Izenda Reports product – stays compliant. Reporting without security is like a car without brakes. You absolutely must have that security. But once that’s in place, our model is about giving open access to the data instead of trying to model every little detail. When you do that, what happens is the users become empowered and consider that their role. IT can then focus on more strategic imperatives.
Are you saying that IT is more a keeper of the data, giving control and access to the end users to do whatever is required by them for reporting?
Sanjay Bhatia: Yes. Data volumes are much bigger than they used to be, and because of compliance requirements, you can’t give someone an Excel file full of data. That laptop could get stolen; there are a lot of issues around that. Plus, the data volumes are so huge – billions of records is the norm now, not the exception. When you have that kind of data volume you have to secure it, and that’s still an IT function. A key is to define the business rules. With Izenda Reports, with a couple lines of code, we can define the business rules, apply the data – for example, salespeople may only see their own clients, regional managers may only see data from their part of the world, division managers may only see data for their employees. But once you defined those basic rules, all the data is wide open. As new data comes in, those business rules are automatically applied, but it doesn’t require an IT process to enable it. It’s all automatic.
Can you tell us why Izenda has chosen to build your platform on Microsoft SQL Server? Do you think it covers the needs of most enterprises today?
Sanjay Bhatia: Absolutely, and we’re very excited about the next generation of SQL Server. With SQL 2012 and Windows 8 Server, you’ll actually be able to achieve 10 TB level databases that don’t require DBAs to optimize the database. You can literally buy a system and over time build up to whatever data level you require. Optimization in the SQL Server world is becoming something that’s not an IT function. It’s really self-tuning; there are wizards now and more of it is automatic, especially as you get into 64-bit, lots of memory and SSDs. We really like the Microsoft platform because our customers can take our tool and plug it in very quickly. Or they can create data stores without having to invest hundreds of hours just to get the databases set up.
What role does the cloud play in data virtualization?
Sanjay Bhatia: The main thing we’re excited about now is the idea of a private cloud. With Windows 8 Server and some of the things that Microsoft is doing, you can use Azure to host your data, and you also are able to have IT be more of a service than a set of people you have to engage. The idea is that a business unit owner can go to the private cloud, and indicate that they’d like a specific application, a specific amount of storage and a certain number of transactions. They can pay for that service, get an SLA and get a refund if there is downtime. They can literally light up applications without having to go through the installation process. You can do that today in the cloud, and that concept is now being brought into private clouds. If you look at the Windows 8 client now, you don’t go through a set-up wizard. You hit one button and you have the application. That’s going to go all the way up to enterprise applications. It’s kind of funny that you can call Dell and they will overnight a server to you, but it will take up to a year to get permission to light it up. With the cloud – public, private and hybrid -– gives you the capability as an application owner to say you need it and are willing to pay for it, but don’t want to hire somebody to make it work.
A major trend that seems to have exploded overnight is social media, Twitter and Facebook data, and sentiment analysis. Are you seeing a need to connect to that data as well?
Sanjay Bhatia: Yes, we are. Those are data feeds. Our clients are very excited about being able to access data on the Azure marketplace or data feeds from sources like Twitter in addition to their internal data. One example is sentiment analysis where you can correlate what’s happening in your business with what’s happening in the stock market. Think about the value of finding correlations between certain commodities or certain stock indexes and the profitability of your business. That level of correlation gives you the ability to actually predict what will happen. Another place where this is really important is healthcare. When you look at certain things like cholesterol, it really involves a lot of guesswork to determine what medication a person should be given or what level of compliance you have on the dosage. But if you look at data for individuals – what they purchase, whether they exercise, what they buy with their credit card and where they live – you can actually do very simple correlations that tell you what’s likely to work. It’s very important for patient populations as well because where somebody lives could affect a clinical trial.
There’s a lot of really cool stuff you can do once you start virtualizing data and incorporating additional data sources. It’s something that really wouldn’t have been possible with a traditional data warehousing mind-set. With data virtualization, as your doing your analysis, you can, for example, use Twitter as a real-time data feed.
How easy is it for someone to work with Izenda?
Sanjay Bhatia: It’s remarkably easy. Typically we will build a 60-second video and a 5-minute video for our customers on how to customize a report. Most people don’t want to build their own reports. What they really want to do is make lots and lots of changes, often involving thousands of reports. Users need to be able to make those changes quickly. With Izenda, you can actually learn how to do that in about 60 seconds.
So someone could be up and running with Izenda in a week? Is that pretty normal for a new Izenda customer?
Sanjay Bhatia: Typically it takes one to two days of work to get it set up and define the business rules and security. If you don’t need security, it’s literally hours.
Is it browser-based?
Sanjay Bhatia: Yes, it’s completely browser-based, HTML 5, works on tablets, smartphones, Android, iOS, and we just announced we’re launching a Windows 8 app as well.
So you’ve got mobile covered as well?
Sanjay Bhatia: Yes, and that’s very critical. When the iPad 2 came out, we almost had to rewrite our product roadmap because we all recognized this was going to be needed for conferences, for boardrooms, for meetings – just the idea of showing somebody a tablet is pretty powerful. The PC has been around for a while, but the concept is really kind of dead. I think you can see that by looking at children. You give them a tablet, and they somehow know intuitively how to use it. That generation is probably never going to want to use PCs.
Originally posted on http://www.b-eye-network.com/view/15939