AI and How It Can Help Get you to the Key Information Faster

by | Oct 30, 2019 | Analytics, eDisclosure, eDiscovery

Artificial Intelligence is the hot buzzword in legal circles but it seems to mean different things to different people, often including Computer Assisted Learning, Technology Assisted Review and more.  

LDM Global’s team recently had a conversation to explain how LDM Global includes AI in its service offerings and to understand more about uses of the technology.

Q. What sets LDM Global apart from other service providers? 

A: From the first client facing meeting, LDM Global acts as an expert. Even our sales consultants are experts. LDM Global sales consultants have been in the industry for years and worked in the industry in a number of roles before moving to the sales side. We even have some former lawyers who are now consultants.  

In my experience at other service providers, the sales side was simply sales focused and didn’t give proper solutions. It wasn’t until clients spoke to the next level of support that the solutions were given. At LDM Global, we step in and give the solutions up front, including getting our project managers and professional services staff involved immediately.  

We are very solutions focused, and we are technology neutral. We don’t just offer clients a product and standard workflow. At LDM Global, we are constantly exploring the market to find the best products for our clients. For example we recently added on a new eDiscovery tool, Nuix Discover. We have also just partnered with company who has a tool that uses AI and machine learning to provide insights into client data. We explore multiple products and give the right product to reviewers.  

Q. With all this new technology, it sounds like it could get confusing for some clients who aren’t used to so much advanced technology. 

A: LDM Global staff are experts in the tools we offer, often holding advanced certifications for which they have spent hundreds of hours training and working to obtain. We therefore can support these tools very efficiently for the client as well as offer various levels of training for the clients depending on what their needs are and how involved they want to get in the tools. 

Q. How can lawyers add efficiencies to their review projects? 

A: There are a lot of different ways we can improve efficiencies for lawyers on their review projects.  

The first step is understanding the scope and strategy of the case. LDM Global project managers work along with the reviewers and the senior associates or partners on the case. Working with the lawyers to understand the case is key so we can determine which technologies would be most useful to implement on the matter. 

You can’t just implement any technology in any case. Technology comes with an understanding of the case and determining which type will add the most value.    

The second step then is implementing that technology at the same time as the review. 

And finally, building the stories in a case can add great efficiencies. For example, we look at whether a litigation is a construction matter or a pharmacy-related case. We can then use what we know about those case types and the correct templates to build the story or narrative for our clients. For example, different types of cases have different keywords, “hot” language and items to look for. By setting up templates in advance based on the case type, the AI system will automatically search and organize the information in a way that shows what we – and it – thinks would be most important or relevant to the attorneys.  We grab all the information and create a story for each person and the way they talk.

For example, I am Murali Baddula. The system may find I have 10 different variations of my name and my email address, which it will pull into one place automatically, but it sees that the way I talk is the same in every email. And maybe I used “excitable language” in some emails, which will be flagged to bring them to the top of the pile so the attorneys can review those first.   

We can create groups of information based on many parameters. For example, if someone was traveling around and taking pictures in different locations, we can build a narrative showing at which point and which country the person was in when he was taking each photo. But with AI, we can do this almost immediately rather than needing to spend hours and days sorting through the information using a standard system.  

Q. Does this mean the lawyers will lose out on billable hours? 

A:  We aren’t taking work from the lawyers in any way. What we are doing is focusing their time by organizing the data into what they want to see first, or what is most important. Technology is not going to replace lawyers on the review front, however lawyers with technology will replace the traditional linear review process. 

The typical eDiscovery workflow has been simply processing documents and putting them into a linear review. What AI is doing is boosting the review speed. They could be reviewing 1 million documents or only 5,000 in a small case. But instead of looking at them linearing, we are reviewing the documents using technology, and our system is automatically learning in the background to determine what is most relevant.  

Q. Are there guidelines on when you recommend using the technology, such as only when you reach a certain number of documents? 

A:  I would recommend implementing it at the start of a case, even if the client isn’t sure whether they will use it. That means basically turning it on and having it run in the background. It will not make any changes to the workflow. But if later the lawyers need to make conflict checks, they could use the technology, for example.  

Once you teach the AI system, it does not think differently. Humans think differently. One document may be relevant to one reviewer but not relevant for the other reviewer. Once you teach the system, it will make the same decision every time, again adding efficiencies to the process.

The system revises every few documents and does active learning in the background so you don’t need a minimum number of documents. For predictive coding, you may need 20,000 to 25,000 documents to start with.  

Q. In your experience is it time saving or cost savings provided for the clients? 

A: Generally, using AI will help reduce both costs and the time of review. You can save up to 70 percent of eDiscovery costs via AI implementation. For example, you don’t need to have 100 reviewers linearly reviewing documents; with Computer Assisted Learning you are only reviewing 5 percent to 10 percent of the document set. So you are basically saving 90 percent of reviewer time.  

You do add costs for the AI technology upfront, but this usually works out to much less than the cost of a typical review had you not used the technology.   

Q. What should clients do if they are interested in learning more about how AI might offer efficiencies in their cases? 

A: Please reach out to your Solutions Consultant or anyone at LDM Global at any time! We love to talk about new technology and ways we have implemented the tools on various cases. We are also happy to provide demos using your own or sample data. You can contact us here: