Seeing Is Believing – and Understanding: FTI’s new Radiance platform visualizes your entire information ecosystem

Saturday, January 30, 2016 - 17:11

JR Jenkins is a member of FTI Technology, the e-discovery and information governance practice within FTI Consulting. Jenkins helps drive product development priorities, particularly in relation to Ringtail, FTI’s e-discovery software platform. Here, he introduces Radiance, a new data analytics tool that can serve as an upstream partner to Ringtail, and highlights its unique visualization capabilities. His remarks have been edited for length and style.


MCC: Tell us, what is Radiance?


Jenkins: Radiance is our brand-new visual analytics software platform that helps organizations make sense of their enterprise data. It’s a highly scalable, flexible platform designed to assist with compliance projects, risk assessments and activities ahead of e-discovery. It allows organizations to connect with, enrich, analyze and visualize a tremendous number of documents, millions of documents, from a wide variety of sources from a single user interface.


MCC: What kind of data can Radiance connect with
and analyze? 


Jenkins: This is one of the most important parts of the application. Corporations today are dealing with well-documented and well-discussed increases in traditional documents – email, business documents, etc. In addition, corporations are rapidly adopting cloud-computing technologies for storage as well as the creation of business information. In our regular day-to-day work on e-discovery matters, we’re seeing an increase in Google Docs and Office 365 file types. 

Google Docs and Office 365 are just the tip of the iceberg, though. Think of the various collaboration applications, structured data software, temporary storage applications – names like Slack, Dropbox, etc. Corporations are using these new cloud-based tools, and often without the central IT team permitting or even knowing about the use. This poses a huge challenge for organizations.

Critically then, Radiance has been designed to connect to the key repositories inside the firewall, the Exchange servers for your email, the file shares where you may be dealing with Word documents, Excel spreadsheets, etc. It can also connect to and then aggregate cloud repositories – the Google Docs, the Office 365 and applications that are dominating headlines today in terms of the business trades, like Slack and Trello. These tools are becoming a primary form of communication amongst teams inside a lot of big organizations.

The ability to bring together not only the traditional output of knowledge workers but also these cutting-edge sources and resources is a critical part of the Radiance story. It’s a primary concern for lot of organizations that may have established workflows on how to preserve, collect and then review some of their email and docs but find themselves scrambling right now as to how to preserve, collect and review Slack data or things in Google Docs. 


MCC: Tell us some of the problems that Radiance is trying to solve.


Jenkins: At a high level, Radiance allows users to quickly look at 100 million documents and make sense of that information. You can see the data in aggregate and pick out trends, or you can hone in on a few documents or interactions between employees. We’re able to do this because of visual analytics. The other day I was thinking about the famous saying “A picture is worth 1,000 words.” A visualization, then, illuminates 1,000 data points. It’s really about trying to provide an environment in which somebody can take a look at everything and narrow down the corpus in a quick and elegant way. It is providing a visual analytics dashboard to your information ecosystem.

As you can imagine, there are a number of needs for this technology, especially if it’s fast to deploy and easy for organizations to use, like Radiance is. As we go to market with Radiance, we’re focusing on a couple of uses: early case assessment, early data assessment and investigation. These continue to be huge challenges for organizations, especially as data volumes continue to grow. With Radiance, you can quickly get to the meaning of this data by helping you isolate and focus on the smaller subsets of your data, such as around specific people, issues, date ranges and other variables. 


MCC: Can you tell us more specifically how Radiance helps with investigations?


Jenkins: The thing that’s most dazzling about the Radiance workflow is its ability to filter down the data in a way that you can go from looking at thousands of custodians and millions of documents to, within just a few clicks, a single custodian, a single person, a single day’s worth of activity, what emails they sent or received that day, and what documents were part of those communications. Investigations are all about trying to identify and isolate a unique set of activities, a pattern of activity. In many cases, investigations start with a couple of facts, a couple of threads, and upon those facts and threads, we weave a more complete story.

Because of this, the Radiance home page is a chronology visualization. It shows a date range along with a list of custodians, and investigators can then very quickly select the timelines they are interested in, filter it even further by selecting one or a few key people and interact with these documents. And with a single click, they can be delving into the social networks and communication patterns of these people. From here, the user can intelligently expand their investigation using the custodians’ own communication as the guide. 

Another important tool within Radiance is concept analysis. During enrichment, Radiance analyzes text and catalogs all concepts – nouns and noun phrases – found in the content. The result is a rich index of “meaning” which the investigator can browse, visualize and analyze quickly. This reduces the burden on the user to know every detail and, instead, accelerates their understanding of the key fact patterns and themes inside the data. We often talk about investigators who are saddled with keywords as their starting point, and where and when keywords are useful. We know that keywords can be used as a way to get you going, but rarely can they be used exclusively to get you across the finish line.

Armed with those small bits of facts – a couple of custodians, maybe a range of dates and a few ideas of what may be at the heart of the issue of the investigation – you can quickly find and isolate critical documents and, in the case of searching across multiple repositories, where they reside. And once identified, Radiance makes it easy to tag documents and ready them for the next phase of the project, which can include export into an e-discovery application, like Ringtail where they can be reviewed with the workflow amongst the team.


MCC: Tell us how Radiance helps with ECA?


Jenkins: ECA, or early case assessment, isn’t a new term for lawyers, litigation support teams or e-discovery professionals. The problem is that as it hasn’t really lived up to its promise. Inside corporations, ECA is, in many cases, seen as a culling exercise. It’s all about eliminating as many documents as you can before review starts in a traditional e-discovery workflow, but very little analysis of the documents takes place. Sure, culling documents is a positive, but what if you can cull and truly do an analysis? With Radiance you can very quickly set aside tremendous volumes of data and do some really rich analysis of the remaining documents, custodians and trends. You can actually get down to looking at the documents. 

How do we do this? One of the really cool things about Radiance is that when documents are brought into the system, there’s a robust enrichment process. That means the documents are extracted and put through a robust text analysis process. This enables us to identify duplicates, near duplicates, threads, file types and size, and a host of other metadata. This metadata is transformed into facets that users leverage as they explore the data set. As an example, language analysis is one of the facets in the enrichment process. A great example of ECA use is that once you have found the 10,000 or 5,000 documents that you’re interested in as part of an early case assessment, you can see how many have foreign languages included. Radiance will tell you not only what types of languages are included in those documents – Portuguese, Spanish, what have you – it will tell you what percentage of those documents are comprised of various languages and even highlight those for you within the document. 

You can very quickly find out what your document volumes are going to be. You can see what the document variety is going to be or the file type variety. You can see what languages are being used. You can prepare much more strategically for a better review because you’ll understand the characteristics of your ecosystem. Plus, Radiance offers a number of dynamic reporting features so this information can be shared at each stage of the project.


MCC: What are some of the visualizations Radiance can offer to users?


Jenkins: The Radiance visualization library includes 10 visualizations that can be used individually or in tandem. They can quickly illuminate millions of data points and support multiple starting points – date, organization, people, issues, languages, etc. – meaning subject matter experts are never at a loss about where to start. 

In addition, Radiance comes with search tools that allow people to leverage traditional keyword approaches, if they have certain ideas that they’re interested in, and you can quickly visualize a search result. 

But we’ve found that investigators find the visualization-first approach truly liberating, as it allows them to interact with the actual document content, not what they think is in the document set.

So our visualizations start with the chronology visualization – it shows the distribution of content across time – as constraining dates is a logical starting point for investigation.

In addition, we have our innovative Document Mapper concept-clustering tool, which helps investigators identify batches of documents that contain similar ideas and terms. We use this in Ringtail for document review and coding, and inside Radiance, the document map and visualizations can be used to look at groups of documents and the concepts that are associated with them. 

Also of critical use is the ability to look at someone’s social network. We know that this is a significant component in a lot of analysis. Whether the investigation is around some notion of internal improprieties or whether it is along the lines of IP, having the ability to analyze somebody’s social network, with whom they are communicating, both inside the domain and outside, with Radiance you can see what the communication patterns look like, what kinds of information are being sent back and forth.

The social network and the ability to look at visualizations across time and see the relationships between people and the information are critical. 


MCC: Tell us how Radiance is different from other software applications that are currently on the market.


Jenkins: From an engineering perspective, Radiance is really focused on high-speed performance and rapid feedback. The software feels nimble even when interacting with hundreds of millions of documents. This was accomplished through very thoughtful and innovative software design decisions that emphasize performance at the database, enrichment and presentation layers. We are keenly focused on visual data analytics at a massive scale, so we were merciless during testing to ensure that any feature or action in the user interface would not slow things down. 

The other thing that makes Radiance different and unique is it is lightweight, meaning that it’s not bolted to something much larger, much more grandiose. One of the things that really stops corporate teams from making investments in the world of big data is the big price tag that comes along with it. We’re talking about multimillion dollar, multiyear contracts that take forever to install and train people on. We know that people want something fast and easy so they can get their heads around issues inside their own ecosystem. 

What we’ve designed here is something that’s very fast, very flexible. It’s designed to solve the set of problems of a big data repository so to speak. And it moves with you into some of the work flows that in many cases are mandated, whether it’s the litigation or legal obligations that come along with being attached to e-discovery or inside compliance and risk offices to identify and isolate the data that may be at most risk.

It’s easy to come in with Radiance, get your data loaded and analyze it without having to make a multimillion dollar investment in consultants who are spending years inside your ecosystem trying to help you understand your data.

MCC: Who should be interested in Radiance?


Jenkins: Those who are involved with compliance, risk and e-discovery should be interested in Radiance. They may be working inside of legal departments, compliance or GRC teams, and even IT. Anyone tasked with investigation. Also, any organization that’s made a move into a cloud environment within the past year should evaluate Radiance for its ability to connect with Office 365 and other tools.

What’s interesting is that if you were in this market three years ago, there was a lot of trepidation about movements to the cloud. There was concern about data breaches, obviously, and the feeling that data inside your firewall may be more secure than data in outside repositories. With some of these incredible investments by Microsoft and Google and others, we’ve seen that what they can offer up in terms of protection is certainly on par with what corporations are offering. 

Now, people have now made that switch over to the Office 365 world, to the Google Doc world, to using Slack, but they don’t have any idea of how they’re going to interact with that data for any of the aforementioned projects. We know those people are concerned about the lack of the workflow around those data types. Those are the primary candidates for taking advantage of Radiance’s capabilities.


MCC: This is another FTI Technology software offering visual analytics to cut through big data. What role do you see visual analytics playing in e-discovery and information governance moving forward?


Jenkins: I think it will continue to grow. I think the simple answer is that visualization is, in many ways, the only way that you can get your head around the document volumes and the variety of data that you’re looking at today. One of the exciting things about the time we live in is that there’s been a real emphasis on charts and graphs and making sense of data. We see it in the use of infographics, which have become such a hot thing. We live in a data-drenched world, and compelling visuals help us make sense of it.

I joined FTI technology almost 10 years ago, and I came to the company because of the visual analytics, because of Document Mapper and its unique visual approach to what was the document review problem of 10 years ago. It is one of the most engaging, innovative and fun uses of visualization that I had ever seen. 

I’m fond of saying that what makes our visualizations different than most is that ours are interactive. They’re dynamic. You are asked to engage with the visualizations, to work with and shape them. While you’re doing that, it’s shaping your understanding, shaping your knowledge of the information you’re working with. At the end of a four-hour review session with the visualization, you have a much deeper connection to the themes, to the people, to the trends that are part of that data than you do if you’re just looking at long static roles, like we see in our email clients.

Every year we’ve had to tell the story about growing data volumes, and it’s true. One of the interesting things that we’re dealing with is not only are data volumes growing but data types are changing. I mentioned cloud-based applications, Office 365, Google Docs. That’s a migration of a traditional data type to a cloud environment. What about Slack? What about Trello? What about some of these other things that we see as part of business communications, part of business intellectual property? What is a Slack post? Is that a document? Is that spreadsheet? How do we account for that in some of our investigations? Visualization of information breaks down all of these types. It continues to give everybody the edge they need when they walk in and need to make sense of the data quickly. 

We’ve seen in e-discovery that the notion of visualization of data has grown in the last few years. It’s no mistake that when we decide to build a brand-new platform that visualizations are at the very heart of it. The visualization experience was first and foremost in our mind. Our users are often subject matter experts who have a deep understanding of their business, about the notion of their unique business activities. They’re not technologists. If you put certain visualizations in front of them, you’re going to overwhelm them. We have a great deal of experience helping nontechnical types, lawyers, make sense of and review massive amounts of data. It’s transforming that metadata and content into recognizable themes and patterns. It was no mistake that our foray into moving left of e-discovery was going to be highly visual. 

That’s the real opportunity for people using Radiance: to make sense of their data much more quickly. Radiance doesn’t impose a structure on the data so much that it allows you to come in and build your own structure around the data, depending on whatever project you’re working with. To do that visually means you’re going to have a quicker understanding, I think, and more confidence in your decision making. As we like to say, there’s this notion of finding facts fast. When you find facts fast, inevitably the strategy becomes your strategy. You own it. You have a much clearer sense of how you want to approach projects versus just reacting to either deadlines or the other side’s demand for information. 


JR Jenkins, Senior director at FTI Consulting based in Seattle.