When our team engineered Relativity Processing—released in 2012—a major goal was to make processing administrators’ work easier. Even against the backdrop of the wider litigation support industry, these experts are often faced with tasks that come with tiny timelines and big demands.
Today, Relativity Processing has been adopted by 138 customers who handle some of the largest jobs in the industry. They’re using the software to tackle terabytes of data for projects such as government second requests, to accelerate their speed to review, and more—and we’re rapidly innovating the tool to help support their success.
A few months after Processing initially launched, we posted a story from the point of view of a processing admin charged with tackling quick turnaround e-discovery projects. Now that we’re approaching Processing’s third birthday (time flies!), here’s another story from the life of a processing admin.
* * * * *
Monday, 8:13 p.m. – Our team just received an email from one of our attorneys about a new project that needs to be made available for review immediately—the case has been fast-tracked for deposition by the court. I thought I was done for the day, but hey—that’s the nature of the job, and at least I can do it from home. There are two custodians involved, and our collection team has already dropped the PSTs in a network share I can access remotely.
I’ll fill out a project kickoff form (PKF) and submit it to the case team. I need to confirm specs for processing the data and setting it up in the workspace—things like relevant date ranges, file sources, custodian names, de-duplication standards, OCR protocol, which time zones we’ll reference, how we’ll handle embedded objects, and maintaining folder structure once the data is in the workspace.
In the meantime, I’m going to start building our workspace so I’m ready to make a few final tweaks and kick off processing as soon as that PKF is approved.
Monday, 8:42 p.m. – The data and our environment are ready to go, so it’s time to get this data into Relativity. I create a new processing set in our Relativity workspace—a quick setup. When I’m done, I specify a folder where the data will eventually be published. The attorney has approved our PKF and requested we set the data to auto-publish once it’s processed. That way, it’s immediately available to reviewers, and my team can perform QC on it to resolve any errors while the review begins. We’ll flag documents with errors so reviewers know they’re being addressed.
Right now, I’m performing an inventory on the data in Relativity ahead of processing. This will give me a high-level view of what’s included and enable me to filter it down to the meatiest material based on the attorney’s guidelines. The case team wants to narrow down the collected data before we drop it into review. The scope of the case is pretty specific, so they’ve provided a date range, email domains, and file types to filter for—information they gathered with legal hold questionnaires.
Once that’s done, I see that we’re dealing with about 60,000 documents from the first custodian and 180,000 from the other. These definitely aren’t tiny data sets to work with. I want to get quick approval of the inventory from the attorney on the case and, since the data sets vary so much in size, I’m also going to see if makes sense to prioritize one over the other to help accelerate our time to review.
Monday, 9:03 p.m. – I just heard back from the attorney. The inventoried data aligns with his parameters, and he agrees we should prioritize the smaller set so some data is available for the review team to start with by morning.
I’m kicking off the processing phase for both data sets now, using my queue manager to prioritize the smaller one so the team can tackle it first. That also means I can run this job overnight—which is great, because it’s almost bedtime.
Tuesday, 6:45 a.m. – I’m up bright and early with a gigantic coffee, ready to see how this job has progressed. It looks like processing is complete and the smaller data set has been published to the Documents tab, so I run the associated reports. I use built-in reports to provide a quick update on the status of the project to our case team. I’m also using sampling and filtering to quickly QC the OCR and images from the processing job. Everything looks good, so my last step is to set up a few standard saved searches to make it easier for the case team to get started.
That done, they can get to work on the first custodian’s data; they have access to the saved searches, and the discovery report I’m sending them paints a picture of the data set and makes it easy for them to start building a plan of attack. We’re reviewing and processing at once here, to help accelerate our timeline.
There’s also an error report for my team, which will help us quickly determine which errors should be prioritized, and ensure we resolve the rest quickly.
Tuesday, 6:01 p.m. – We’ve resolved most of the errors on the first data set, and the review is running smoothly. I just got an email notification that the second custodian’s data has finished processing as well, so I’m going to repeat the QC, reporting, and searching setup I did for the first one while they add that data to their review plan.
It’s been less than 12 hours since we received the initial request, and I’m happy to see the case team already reviewing data while my team tackles some minor cleanup. That way, we can both execute on what we’re best at simultaneously and keep this case moving along.
Wednesday, 2:44 p.m. – All of the errors on both of those processing sets are resolved, and the review team is on track to tackle the data by their deadline. I breathe a sigh of relief and check my email; that’s the good news. Not so great? Here’s a note saying I should expect 200 GBs of data that need to be immediately available for review in another ongoing case…