Facebook

Brian Hupp, head of Legal Operations at Facebook, can’t help it if he sees data as a burden. Sure, the data Facebook hosts and creates is at the core of its business strategy and competitiveness. But for Hupp’s team, the volumes of data to manage are growing exponentially, and being hosted in more repositories than they can count.

“Data sources tend to multiply like rabbits,” Hupp said Tuesday at the “Knowledge Management + AI: Quick Wins Through a Connected Ecosystem” session at the 2017 Corporate Legal Operations Consortium’s annual institute in Las Vegas. “Not only are your legal team members creating new data sources for themselves, a lot of them still like to work in the old way, creating their own Dropbox repositories, saving things onto their C-drives, saving things onto shared folders, onto Box, et cetera.”

And what’s more troublesome is that taming this data will likely never be a one-off effort. “Even when you get a handle on the initial one [data source], then your M&A team will come along and buy another company, then you have another set of systems,” Hupp explained.

He added that the problem of disparate data is also more acute in corporate legal departments than other in-house teams because different practice groups “literally sitting 10 feet away from [each other] can have their own data management plan, their own data repositories that they are using.”

Though it is challenging, having visibility into all this legal data is paramount not only for efficiency and cost-saving purposes, but more importantly, to meet regulatory demands.

A lot of compliance is “knowing what’s in your data,” according to Hupp. But in using disparate data repositories, he said, “even if you think you’re [not storing or managing] regulated data,” you can often be wrong.

To tackle this dilemma of disparate data head-on, Hupp said the Facebook legal operations team looked to “bring the benefits of e-discovery search into the department” to meet a host of objectives. These included connecting data sources across multiple systems, ensuring any data management effort wasn’t itself too burdensome on the legal operations staff, overlaying workflow and notifications on top of in-house data, and finally enabling quick and easy browsing of all data the department held.

The “old school way” to accomplish that, Hupp noted, “was to hire an extensive team of outsourced workers to look at [and organize] every single document.” But now, Hupp’s legal operations team deploys the artificial intelligence-based knowledge management platform from Onna, which uses machine-learning neural networks, to automate the entire process.

While automation like this may have once seemed far-fetched, Salim Elkhou, founder and CEO at Onna Technologies, who was also a speaker on Tuesday’s panel, said such a task “is really not that difficult” for modern AI e-discovery and information management platforms on the market today.

“A lot of these providers have very strong application program interfaces (APIs)” that can connect with data sources such as Gmail or Dropbox to analyze, index and organize all the disparate stored information, Elkhou said.

So far, Hupp’s department has been impressed with the “speed through which you can integrate a lot of these disparate data sources,” by using Onna’s platform.  One of the first things he said his team used the platform for was to organize data “on an individual level, as each of us individually have a lot of data in various places—we are inundated with it.”

Leveraging AI has allowed Hupp’s team to broadly or specifically search and filter information they hold to respond to a variety of legal and business demands. “The quick wins for the team are really being able to share the data [they hold] across a topic, a policy, whatever. If they need [certain data], they can get it really quick on their own with almost no hand-holding,” Hupp said.

Contact Rhys Dipshan at rdipshan@alm.com. On Twitter: @R_Dipshan.