E-Discovery and the Cloud

Greg Arnette, founder and CTO of Sonian, strikes me as a serious and studious guy, who understands the deep implications of his business. I like that. His business, just so you know, is creating the technology and infrastructure that supports e-discovery, especially in the cloud. He's deeply into the cloud.

The forces of change for Greg include the impact of cloud computing, the exponential amount of data creation, and the degree to which it is a new burden on IT to wrangle it all.

Now think about the IT organizations in most companies. Multiply "IT departments" by "data to be stored in the cloud," and you have a great shifting of thinking in how organizations decide the best policy to store data, to discard with confidence that data, and to be prepared to assemble the data that might be required to react to a specific event.

And then there is the impact of the various vertical markets that the organization might be part of, or states, or localities. It's often a geo-political environment-there are many different places and rules regarding where data can live. And then there's the recent news emerging surrounding the NSA, regarding who can see what... It's extremely complex. Whew.

Yet, Greg insists these challenges can be met, and be met relatively inexpensively. Which is good news. He insists there is a way to store, encrypt and make searchable data in a cloud environment that allows rapid and accurate e-discovery across hundreds of content types and large stores of "variety-data," I call it. Stuff you didn't have only a few years ago. It's a new world of text plus database plus email plus...who knows what. That "who knows what" is the basis of the new information paradigm. No one knows "who knows what." That's the challenge for IT today, and also for the business managers who need ready access to those data that support their decisions and create opportunities for value to their organizations.

"We have demonstrated that the cloud is a viable place for sensitive, long-term storage of employee-generated content, and there are metrics to prove that this is the right path." That's a bold statement, coming as it does in the era of not-so-slight skepticism of cloud computing and its relative result on discovery and security. But Greg is adamant. And convincing.

"The buyers of the service are non-technical people. They are running HR, compliance, the C-suite... they are the purchasers, but they direct their IT people to find the solution to find the capabilities." That's fairly common, I reason, and agree.

"But," Greg admits, "there are two momentums happening. There's an unhappiness with the on-prem systems that have been doing this for a decade, and then there are the people who have been priced out because on-prem systems are so expensive. Six-figure, seven-figure systems were so expensive, companies said, ‘We'll just have to wing that,'" Greg says. "So cloud economics come into play."

Low Risk, High Reward

Greg thinks that new companies have been priced out. On-prem solutions are too expensive. Cloud solutions, he thinks, are attractive, but somewhat spooky for large organizations. There's probably a little of each.

"That's the way it works. You only pay for what you use," says Greg. And he's right in most ways. "Whenever there's economy of scale that trickles down the stack, the IT people see that, and they're being directed down the stack also."

But my question is whether IT is welcoming or resistant to the cloud. I wonder whether IT departments withhold a powerbase, and letting their control over on-prem systems represents a release of that power? Or whether it means extra "brownie points" with management that gets them extra love over time because of their devotion to the cause of bringing lower costs and better service?

"We don't try to wholesale move everything to the cloud," says Greg. "We don't do that. Those kinds of projects put too much fear onto a traditional IT person." But Greg assures me that certain projects can retain the same user interface that they're accustomed to. And they do not need to feel threatened, job-wise.

"There are lots of data workloads that will need to live on-premise for a long time," predicts Greg. "Things that require low-latency access, real-time systems, things like that don't need to go into the cloud right away," says Greg, in a soft moment. "But a system that is designed to incrementally collect data all the time, 24/7, build search indexes, include high levels of encryption, and make it searchable on demand... that's a hard thing to do on premise. That's the perfect cloud use-case."

"But," he adds, "there's a new wave and a new guard. The on-prem thing costs a lot and didn't always help."

Moving Upstairs

The C-suite is gaining more education and are becoming more aware of the technology. That is a good thing. They know the vast difference between capital expenditure and operational expenditures, and this area of document management and content management matters a great deal.

"As usual, it's a combination of things, as a lot of things are," says Greg, wisely. "There's more awareness, and more people planning the policy stuff. We're finding that customers are getting more comfortable with not just keeping stuff, but also deleting with confidence.

"People need to be trained on policies, and then defend the policies, but then you need the technology in place to do the ‘trust but verify.' Technology is getting to the point where you can have ‘sensors' or ‘agents' that make sure all the right data ends up in the sanctioned source or truth. It feels a little ‘Big Brotherish,' but if it's important to that organization, that's the ‘trust but verify' component."

It will be a significant matter for many organizations. Greg thinks there is a population that will embrace it slowly, and a population that will like the "attractive pricing and the easy way to get up and running."

And then there's the rest. Those have been a dampening factor. Preparing for e-discovery is a little like buying an insurance policy. You might not need it. Or you might. These days... you probably will.

I'm not in any position to make that call. But most of you are, so I recommend you read the rest of this white paper and make the best judgment possible.