Virtualization Technology News and Information
VMblog Expert Interview: Scott Lucas of Concentric Talks Securing Unstructured Data and Semantic Intelligence


Concentric has launched some significant advancements to its flagship data access governance solution, called Semantic Intelligence.  To learn more about the company, its latest offering, and securing unstructured data, VMblog spoke with Scott Lucas, head of marketing at Concentric.

VMblog:  This is our first conversation with Concentric.  Can you please give me a quick background on the company?

Scott Lucas:  I would be happy to Dave. Concentric was founded in 2018 out of a powerful insight: Enterprises have a data security problem. And the security tools they're using can't fix it.

We're a team of AI experts, data scientists, software builders, and security junkies dedicated to solving the most pressing data security problems. We live to serve our customers and solve their toughest issues by delivering the latest technologies. We won't rest until the last contract, design doc, trading sheet, strategy plan, and M&A terms sheet is safe and secure.

Concentric mitigates risk to sensitive data by autonomously identifying business-critical documents that are inappropriately shared or incorrectly classified. We isolate the highest risk documents out of the millions of contracts, source code files, personnel documents, and strategic plans in your organization.

VMblog:  Tell me about this new solution you are launching to protect unstructured data.

Lucas:  We are launching some significant advancements to our flagship data access governance solution - called Semantic Intelligence - that address the industry's continued challenges with unstructured data security threats. We are the first in the industry to autonomously identify risk created by inappropriate access, sharing and management of unstructured data. This includes on-premises, cloud storage and cloud productivity application environments.

Concentric's Semantic Intelligence solution now offers patent-pending Risk Distance analysis for autonomous assessment of file and activity-based risk. Risk Distance uniquely uses advanced deep learning technology to find at-risk data by comparing each file to its peers. These peer file comparisons give security professionals, for the first time ever, the highly targeted and actionable guidance they need to finally be able to secure their business-critical unstructured data without hard-to-maintain rules or complex configurations.

VMblog:  Why is it so difficult to protect this type of data?

Lucas:  Concentric is tackling a hard problem - securing customers' diverse, dispersed and highly specialized unstructured data which, according to Gartner, represents 80 percent of all corporate data. Protecting millions of disparate files stored across a sprawl of data storage options has proven to be an impossible task for IT professionals.

Previously, identifying and assessing risk in unstructured data was a guessing game. End users control security-critical decisions, and available tools can't differentiate between files that are business-critical and those that aren't. As a result, prior attempts to address this ongoing problem offered only limited insight and incomplete protection. In fact, recent data security incidents at a major bank and drug manufacturer highlight how even large, well-funded organizations cannot fully secure their unstructured data.

Work from home practices are making the job tougher too - and that's a trend that appears here to stay. Remote teams use links, personal email and the cloud to share data and get work done - often without a thought to the security consequences. This puts sensitive data - such as intellectual property, business strategies, and PII - at risk.

VMblog:  Aren't there already solutions out there addressing this problem?

Lucas:  I think it's fair to say there are tools but no real solutions. One approach relies on pattern matching to find business-critical data. You might, for example, scan files looking for text that matches a pattern for government ID numbers. IT teams are on the hook to develop and manage these patterns (or "rules"), and that has led to a maintenance nightmare at every organization who has tried it. Accuracy's a problem too. You can't expect an IT team to have the specialized knowledge needed to understand document content, which means they can't write very effective rules. So you end up with an inaccurate, incomplete solution that needs lots of care and feeding. Not an ideal outcome for sure.

Because of all the problems associated with rules-based approaches, organizations now often lean on users to classify their own documents. If you've used Microsoft AIP you've probably experienced this firsthand. It's up to you as a content owner to mark your documents as sensitive or confidential or whatever category applies. This avoids the messiness of rules but opens up a host of other issues, such as training and user adoption. Also not an ideal outcome.

Concentric's deep learning solution automates the problem to avoid rules, training requirements and end user involvement. It's the industry's first effective alternative, and it's the only way to accurately identify and control the sprawl of unstructured data.

VMblog:  Tell me how you're using artificial intelligence in your solution, and how does that benefit customers?

Lucas:  To be specific, we use a type of artificial intelligence called "deep learning," and we use a branch of deep learning call natural language processing (NLP). NLP excels at textual analysis (as opposed to image recognition or other tasks where deep learning's been successful). I'll refer to AI here as "NLP".

NLP is a part of our solution in three major ways. First, we use NLP to organize files into one of more than 90 standard categories (customer-defined models are also supported). This is critical because you have to know what you have - and whether it's business-critical - before you can work on risk. NLP provides the semantic understanding of your data that then becomes the foundation for everything else.

Second, NLP powers our Risk DistanceTM analysis, which compares the security practices of an individual file (where it's stored, how it's been shared, etc.) to the practices followed by that file's peers. For example, if only a handful of files in a category are stored in a folder accessible by everyone, Risk Distance will identify that handful as high risk. We also look at file activity, such as file downloads, copying, printing and so on, to get a sense of risk urgency. Risk Distance shows you the intersection of business criticality, security anomalies and urgency, which is a very complete picture of business risk.

Finally, we offer a capability called Concentric MIND (part of Semantic Intelligence) which is a deep-learning-as-a-service infrastructure that curates all available models (developed by Concentric or customers) to offer the best model option for a given data set, helping extend the solution to specialized applications and new markets.

We automate the entire process, from data discovery and classification to risk identification and mitigation. The customer experience is a dramatic departure from rules-based or end user classification alternatives. No more rules to maintain and no involvement - at all - by end users.

VMblog:  What are some of the key benefits organizations can realize with your new Semantic Intelligence solution?

Lucas:  Let me get out of the AI weeds for a moment and refocus on Concentric's mission, which is to secure unstructured data. Today, content owners are making most security decisions and they aren't always thinking about security when they link-share (or otherwise grant access to) a file.

When you limit data access to only those who need it, you're implementing a "zero trust" security model. Zero trust is hard for unstructured data. You can put all kinds of controls on a database but it's far more difficult to do that on files created and owned by end users.

Concentric enables zero-trust for unstructured data, autonomously and without end-user involvement. It's that simple.

VMblog:  Can you describe a typical customer for this new solution?

Lucas:  Any organization that has or handles sensitive and private data can benefit from Concentric. We have seen significant traction in financial services and healthcare, which regulate data privacy, and in other markets such as high tech and manufacturing, where issues like intellectual property protection and competitive advantage are perhaps bigger motivators.

VMblog:  And I can't let you go without asking, what can we expect to see from Concentric in 2021?

Lucas:  We will continue to bring the latest and most effective AI capabilities to market. It's an exciting and dynamic technology - accuracy and performance are improving all the time - and we're committed to bringing the best technology to bear on data security. We have a few other surprises in store - but I'd hate to spoil it for you!

VMblog:  It has been great speaking with you, Scott.  Anything you want to add or leave our readers with before we wrap up?

Lucas:  Concentric's the first new thing in the data security market in years. If you're looking for an effective solution to protect your business-critical files and documents, you should give us a call.


Published Friday, November 13, 2020 7:37 AM by David Marshall
Filed under: ,
There are no comments for this post.
To post a comment, you must be a registered user. Registration is free and easy! Sign up now!
<November 2020>