The refrain “It’s people” may bring back memories of the 1970s dystopian thriller "Soylent Green" for some people, but you can be assured that in ediscovery it is nothing quite so nefarious. The reality is that even with the best technology, the largest computational lift, and the biggest, baddest AI models, you have lost the race before even leaving the starting gate if you do not have people who know how to effectively use it.
Don’t worry though, this is not a call to learn to code or worse yet how to do math (perhaps I am projecting a bit here). Instead, this article takes a look at the importance of intuitive and easy-to-learn technology that even the most technophobic member of the case team can not only learn to use, but can master. I’ll also highlight the benefits your team can achieve using a supercharged solution like DISCO.
Why people matter
More robust technology, advanced processes, and turbocharged AI are all part of the equation for winning the ediscovery race. However, these amplified results are contingent on case teams and document review teams easily understanding how to engage with the review platform and underlying AI without running for the hills. There is no easy button in ediscovery (at least not yet). The benefits of faster tech and more intelligent AI models are contingent upon humans understanding how to interact with the review platform to quickly gain insights. If the review team can’t or won’t leverage a platform to accelerate speed to evidence, even the most advanced review platform and AI models will not yield results any better than old legacy technologies.
How DISCO takes everyone along for the ride
Thankfully it does not take the ability to code, an advanced degree in computer science, or an understanding of lambda calculus for practitioners (technophiles and technophobes) to get the most out of DISCO’s supercharged speed.
Baked-in AI makes you smarter
On any day, most of us interact with a half dozen or more forms of AI before we have a mid-morning coffee break. From Facebook’s “people you may know” to Netflix’s recommendation queue and conversations with Amazon’s Alexa, to less obvious instances of AI like suggested routes on your GPS or targeted advertising when you google something. Every day, in dozens of ways, AI is seamlessly integrated into life without drawing attention to the complex computations and algorithms behind it.
Historically, legal technology took the polar opposite approach. Rather than making AI integration seamless, legacy tech required an alternative workflow, linguists, and statisticians to use AI. In many cases, an entirely new software license (for a sizable fee) was required to leverage AI at all. In the event you wanted to embark with AI, data had to be processed by this new software and based in the tools.
DISCO thought this approach was nonsense. DISCO AI is activated at no charge for any case, with the simple toggle of a switch from off to on in the AI tab of the platform. The AI runs in the background and learns based on each coding decision. After just a few dozen decisions, users start getting better and better insights. Naturally, reviewing in context promotes better decision making on similar documents and case teams are able to dramatically speed up review even if they make no change to their workflows. The net result is that more case teams get better results with easier-to-use AI in DISCO.
What does DISCO’s AI-supercharged managed review look like in practice? On day one of a review, the case team found that 2% of docs were responsive. After DISCO AI took inputs from a single day of reviewers making coding decisions, 79% of what was reviewed was tagged as responsive. After a single day of interacting with DISCO AI, our reviewers went from just one out of 50 documents suggested by AI being responsive to 4 good learning experiences in every 5 documents.
Smart batching makes you smarter
DISCO automatically assigns batches as the users need them, pulling data that reflects the current best documents based on DISCO AI. Why does this little thing make case teams smarter? First, the case team always has the best model applied to their batch selection and benefits from all of the insights from the case team in near real-time. Secondly, because we are batching off of a common tag and prioritizing documents, we know that our reviewers are working with similar content. As such, tagging behavior should be very similar across the team. The platform analyzes and displays tag usage so that we can easily identify outliers. When we see this, we examine outlying batches and revise coding where needed.
This consistent and informed approach is the DISCO approach to improved accuracy. An old John Deere advertisement sums this up beautifully: It’s not how fast you mow; it’s how well you mow fast.
DISCO is driven
The DISCO Review team are Grand Prix level drivers of our platform and as such, are able to review documents faster than the industry average. Adams & Reese LLP was representing a client in a large commercial dispute involving expedited discovery, and had one week to make a production from thousands of documents. A review team of three DISCO power users delivered speeds of 86 docs per hour with highly accurate results. With DISCO AI under the hood, the team found 86% of the responsive documents within three days after reviewing only 48% of the document population. And if that victory wasn’t enough, the managed review team completed the review ahead of schedule and under budget.
Most of the industry relies on random sampling and targeted searching to identify mistakes and ensure accuracy. This approach generally yields a 20% rereview rate for QC. On a 100K document review, this means reviewing 100K plus a rereview of 20K docs, sometimes at a higher hourly rate. DISCO, by contrast, uses a bimodal approach that weights intrateam alignment and deviation from AI model suggestions to flag high-priority candidates for QC. This approach combined with a smaller volume of random sampling and targeted searching generally is able to fully QC a review set with only 4% of the documents being rereviewed. On the 100K document population, we may rereview only 4K documents to ascertain accuracy — 16K fewer documents than legacy approaches
How can we do so much less QC? The AI makes predictions not just for the purposes of prioritization, but also for QC purposes. When a reviewer codes a doc in opposition to a prediction, the document is flagged for rereview in real time. Case teams don’t have to spend as much effort looking for the mistakes, because the mistakes announce themselves. Flagging potential errors also enables a tight feedback loop. Continuous monitoring and rigorous coaching mean that many mistakes are avoided.
Netflix-simple to use
Netflix is so easy my five-year-old niece and 80-year-old grandfather can just as easily interact with the queue to get the type of movies they like. DISCO took a similar approach throughout the platform, making a super powerful and bleeding-edge technological backbone supporting intuitive solutions to your ediscovery problems approachable and far from intimidating. Take data ingestion for example. Rather than having to stage data, remove it from any containers and then make dozens of discrete decisions about processing DISCO created a High-Speed Uploader that is even more powerful than legacy tech, but drag-and-drop simple to engage with.
Screens that were complex to navigate and contained enough individual processing decisions to give the most veteran discovery experts a migraine were replaced with simplified wizard screens that walk users through just a handful of processing choices before kicking the process off. Whether you are adding a gig of data to an existing case or kicking off a 14 TB matter that spans 6 countries, this intuitive user interface allows practitioners to kick off the process without breaking a sweat.
This combination of powerful tech and intuitive UI is evident throughout the entire DISCO platform, from AI that does not require a complex workflow and is turned on with a simple switch to easy-to-use data visualization that empowers even a tech layman to conduct data gap analysis, social networking analysis, and content prioritization with a quick glance. Every step of the way, DISCO empowers users of any level of technology fluency to make impactful insights and amplify their decisions throughout the data set.
Compared to the training rounds, statistical validation, and manual Excel export of legacy tech claiming to offer model-sharing, DISCO’s integrated and user-friendly approach to AI is a game-changer. Users simply have to toggle on AI or cross-matter AI by model topic to begin getting the results. No alternate workflow, no complex data ingestions and staging in a middleware tool — just click and get better insights. This ease of use combined with a price tag of free for DISCO AI takes away many of the barriers I faced when dealing with case teams. By simply engaging with DISCO AI, case teams and review teams become better ediscovery drivers.
Did I mention it's free?
Unlike legacy tech which can cost between $50-150 per GB or possibly require acquiring new servers and licenses, DISCO includes AI and cross-matter AI on every case for every customer at no charge. The belief is that cases using AI get better insights for less time and money, we want to remove any hurdles from using it in every case. The ability to easily use all the bells and whistles in and of itself makes DISCO users better ediscovery drivers.
Undertaking an ediscovery matter with DISCO is like standing at the starting line of a race with the fastest, most-cutting edge car, facing a shorter race track than the competition, and with the confidence that you can be (or have) the best driver. DISCO is specifically designed to be intuitive to people of varying levels of technology fluency — so the entire case team can become DISCO experts in no time. Whether you are looking to drive DISCO yourself or want to engage with our legion of DISCO managed review experts, you can be certain that DISCO will help get you across the finish line.