Unstructured Legal Data and the AI Lifeboat Solution

The world is drowning in data. Music data. Photo data. Video data. Spreadsheet, flow chart, email, and website data. These days, no matter where you are or what you’re doing, there’s someone nearby creating data—even if it’s just CCTV footage or the satellites orbiting our planet. Even you are creating data right now as you browse this article, creating a digital footprint that is being recorded in a big bank of computers somewhere.

Data isn’t a bad thing. It’s just…a thing, a product of our life in this modern era, where, even if you aren’t actively creating it yourself, someone nearby is probably creating it about you.

For modern attorneys, this digital information is the lifeblood of a case—especially when it comes to unstructured data since electronically stored information often contains vital intel that is critical to success. However, the sheer volumes of it are crushing attorneys, making it nearly impossible for legal teams to meet important discovery deadlines and keep client costs down, while at the same time, staying sane.

Fortunately, AI technology can help legal teams optimize workflows and better manage this ever-growing mountain of information.

With the introduction of machine learning software, attorneys now have full control and command over all types of data evidence (instead of the other way around). These data strategy solutions enable legal professionals around the globe to meet modern problems with modern solutions.

Here’s what you need to know about structured vs. unstructured data, how both are affecting the legal industry, and what Veritone is doing to help shape the AI legal revolution, providing dynamic solutions to the pressing dilemmas faced in e-discovery, today.

Structured vs. Unstructured Data: What’s the Difference?

The world might be drowning in digital data, but in the digital world, not all data is created equal. Indeed, when it comes to electronically stored information (ESI), data actually comes in two forms: structured and unstructured.

Structured data exists within the confines of an already established database (for example, Excel). This material is fairly simple to search and categorize, since it already has that built-in support system, with clearly defined patterns and parameters.

Unstructured data, on the other hand, is more difficult to search, since it lacks the internal skeleton that structured data comes pre-equipped with. This type of data basically encompasses every other type of media you can think of, including emails, text messaging, office memos, videos, photos, CCTV footage, jailhouse recordings, and so much more.

Both structured and unstructured data appear often during document review, and each has the potential to contain vital, case-building insights that attorneys need to be successful. Unfortunately for contract attorneys, however, it just so happens that unstructured data (the “harder to search” variety) makes up the bulk of most e-discovery projects—and not by a little, either.

Unfortunately for contract attorneys, however, it just so happens that unstructured data (the “harder to search” variety) makes up the bulk of most e-discovery projects—and not by a little, either. According to Juliette Rizkallah, CMO of Kong, Inc., a whopping 80% of electronically stored information that’s reviewed during e-discovery is unstructured, and that was back in 2017.

Here’s a closer look at these two types of data, and how they’re each affecting legal document review.

Structured Data

First off, let’s take structured data—the easy, “golden child” of electronically stored information.

This type of data typically resides in a relational database management system software (RDBMS), which essentially means that the data is organized and stored in tables, like a spreadsheet. This information is generally short and succinct and can be anything from zip codes, to social security numbers, to payments, and even names—basically whatever can fit neatly into a row-and-column-type organization system.

However, relational databases are so much more than just a convenient method of storage and organization. As anyone who has used Excel to execute payroll, keep track of projects, or compare yearly earnings knows, the power of RDBMS goes beyond storing information; it can also compare and contrast that information with the data stored in other relational databases.

What all of this adds up to, is a body of information that is specifically designed to make it easy for users to organize, search, filter, categorize, and compare, making it quite simple to navigate during electronic discovery.

And then there’s unstructured data.

Unstructured Data

On the other side of the data tide pool, is document review’s problem child: unstructured data, which is responsible for creating at least 90% of all e-discovery headaches (give or take).

One of the main reasons unstructured legal data is so problematic is because there’s just so much of it. Unstructured information can be either human-generated or machine-generated, and it makes up the bulk of information we’re used to encountering on a daily basis.

Here are some of the most common examples of human-generated and machine-generated unstructured data:

While these lists aren’t exhaustive, they at least give a good general sense of what types of information fall under unstructured data. And when you consider how much is being produced, daily—not just by the singular micro-influencer on TikTok—but by global corporations around the world, the amount is staggering.

This brings us to the second problem with unstructured information: its lack of organization.

Unlike its structured counterpart, unstructured ESI doesn’t follow any kind of organizational rhyme or reason. As its very name implies, this type of data is chaos. It is generated by countless sources—in at least as many formats—and even when it straddles the line of being semi-structured (such as the case with email or IM chats), things are still messy, making it a nightmare to collect, review, filter, and catalog during electronic discovery.

Luckily for those still reeling from this one-two punch of volume and chaos, Veritone has a solution: aiWARE.

With its unique ecosystem of cognitive search platforms, aiWARE now gives attorneys the power to structure the unstructured faster and more effectively than ever before.

Veritone aiWARE: A Unified Solution to Unstructured Legal Data

Because unstructured legal data has so many variables and unsearchable moving parts, Veritone set out to create a unifying solution: They envisioned a data strategy that could quickly and easily locate any information a firm wanted, without needing an A.I. expert on call. It was under this premise that aiWARE was born.

Veritone aiWARE is the first operating system of its kind to be designed specifically for artificial intelligence. Rather than requiring users to pick and choose from among their favorite platforms, aiWARE joins hundreds of the best A.I. machine learning models into a single, easy-to-use interface.

This unification is groundbreaking. It means that firms no longer have to be selective about which A.I. software they want to deploy on any given project. Instead, they have the power and variability of hundreds of different programs, allowing them to customize search parameters and receive actionable intelligence at near real-time speeds.

And the best part is that the benefits of these powerful search and organizational tools are not limited to text documents alone.

Objects and Images: Search and Destroy

Unstructured data isn’t limited to text information, so why should its solution be?

Veritone’s AI solutions for legal professionals offer the ability to expand search parameters into the realm of videos and photos as well as text. With Veritone Evidence, a platform of customizable tools including Veritone iDentify and Veritone Illuminate, at their disposal, attorneys can scan unstructured media for faces, objects, logos, and so much more.

With the information identified, Veritone Redact can then be implemented to eliminate sensitive and personal information from the records, helping to blur out faces, redact license plate numbers, and remove other identifying personal information from film and photo faster and more accurately than ever before.

These tools enable firms to effectively meet compliance requirements, while also protecting sensitive information from the public eye. In addition, they have the added perk of granting attorneys valuable, early case insights.

Veritone Illuminate: Establishing Early Case Insights

As if the ability to search images and videos weren’t already great enough, with Veritone Illuminate—an early case assessment addition to the aiWARE universe—attorneys can now transform audio, video, text, and other unstructured media data into early case insights.

This early case assessment helps attorneys identify key issues in a case, highlighting strengths and weaknesses, identifying important search terms, quantifying the amount of relevant data, as well as citing potential costs and liabilities that may arise. These insights can then be used to narrow the focus of a budding lawsuit, ultimately benefiting both clients and legal professionals, alike, in the untold number of hours and dollars saved, overall.

Veritone’s Worldwide Connection

Veritone might be a US-born company, but unstructured data isn’t as particular to nationality as we humans might be—and neither are lawsuits. This means that in order to be effective, data strategy solutions cannot be limited to our own, backyard problem-solving.

Instead, as our horizons continue to broaden, our AI solutions must broaden with it. That’s why Veritone solutions are not limited to ESI found in English alone. Instead, these AI-powered tools give legal professionals the ability to not only transcribe unstructured data, but to translate it as well, fully equipping them to meet the global phenomenon of unstructured legal data in whatever language it might present itself in.

To learn more about Veritone’s AI solutions for legal professionals, contact a team member today.