The following interview was conducted by Doug Austin of eDiscovery Today.
Thought Leader Interview with Dr. Gavin Manes of Avansic: eDiscovery Trends and Best Practices
I recently interviewed Dr. Gavin Manes, CEO of Avansic. We covered so much with regard to eDiscovery trends that we couldn’t fit it all in a single blog post. Part One of my interview was published Monday, part two was published Wednesday, here is part three with Dr. Gavin Manes.
In part three with Dr. Gavin Manes., we concluded the interview by discussing the acceptance of AI and TAR, alternative pricing models for eDiscovery and what Gavin and Avansic are working on! :-).
Doug Austin: There's a lot of discussion about artificial intelligence (AI) and technology assisted review (TAR) being applied to eDiscovery workflows, but a lot of legal professionals still haven't embraced the use of these technologies. Why do you think that is and what can be done to change it?
Dr. Gavin Manes: I think expense is still the number one cause for not using analytics and AI tools. Some vendors might charge extra for access to analytic tools. Some vendors might charge hours to help you with that tool. Some software companies require an extra license or requires an extra server. I think that's been one of the main barriers to people using AI more.
Now, in our practice, every case that goes into the review platform has some type of AI run on it, whether it's “near dupe” or clustering or “find similar”. Whether that client is going to use something like TAR or CAL is another level where they're not trusting the technology, just like we discussed earlier. They've now embraced Zoom and understanding how it can benefit their practice. It's about trusting the technology. To do that, they need a win, they need to realize how it actually helps.
We've had so many cases where just using simple “near dupe” technology removed a huge review burden – 25 to 30 percent of the data. That’s not really AI to the point where it's making decision for you, but it's determining that, “I produced that document already. They produced it back to me. It's the same, so I don't need to review it again.” Or “Here's a document that I produced that I knew was hot. They didn't produce it back to me. Why is that?” Analytics and AI are far beyond what we typically think of in terms of TAR and CAL and clustering. It's all these other analytics that we can do to help us apply automation where we're not having to have a human go and look for things.
Now, when you get out to TAR and CAL, there is a bit of a learning curve and a trust system out there, but that's where we as experts can help you through that path. I’ve seen more non-traditional use of CAL – not from a perspective of having to review all this data and produce it – but instead from a perspective of receiving all this data and needing to review it and find the “hot” documents. And CAL is helping us get through there a lot quicker for people's depositions.
One of the common things we did in the oil spill case where we had four reviewers that were reviewing manually the old way (using search terms, finding documents), was that when they would find a “hot doc”, they would win a steak dinner. And we, as the “nerds” would then take that “hot doc” and ask the system to “find more like this” and we would find a handful and submit it to the team and they would decide those original hot documents were “garbage”, but these are prime. What would we do then? Take those prime documents and do that regression again until we get the best stuff out there.
It's just like you're shopping online. You're looking for a red shirt. It shows you a blue one and you decide “hey, I like that blue one. I'm going to click on that.” And you go down that world. The difference is the online shopping algorithms are not trained by humans; they’re trained by the system that was built out there.
So, the barriers I’ve seen are that it’s hard to understand, it’s too complicated and I haven’t had success with it yet. Those are the barriers that I see in addition to the price barrier, which I hope just goes away across the industry.
Doug Austin: That’s a great segue into this next question. We've seen increased recent discussion regarding alternative pricing models for eDiscovery services and use of cloud-based eDiscovery solutions. Do you think the market is shifting regarding how services are priced and what alternative pricing models do you predict will become more common?
Dr. Gavin Manes: I think it goes back to something we talked about previously. This DIY concept is changing pricing where we have a “white glove” service where I will do everything for you and help you. On the opposite end, we have a subscription service where the system is, “Go away. Call me if you need me.” I think everybody lives somewhere in the middle and you shift back and forth based on the needs of the case. So, as we shift more towards DIY and everybody gets comfortable, I think how that’s priced is going to change.
We mentioned the old Clearwell model, if people remember that way back in the day. You had a bucket of data and you could fill that bucket up, use all this data in there, throw everything out of the bucket and put more data in. You were never charged for what went in the bucket, you were only charged for the size of the bucket. It's “all you can eat ice cream” as long as you can keep getting ice cream in there somehow. I think the model that's probably going to reign supreme is one where you buy buckets of storage.
It's very analogous to the way we buy online email. If you're in Gmail or Office 365, you're not being charged for every email that comes in a system, you're being charged based on how much data you have stored at one given time – not the unique data. I don't think that's going to take tectonic plates to move. I think once the insurance companies behind the scenes realize there's a cheaper way to do eDiscovery, that realization will push the industry out there.
That pushes you towards subscription services where you're not paying per gigabyte per month. Because with that bucket, if put a gigabyte in and I then produce half that gigabyte, I'm filling my bucket up, even though I haven't put any new gigabytes in, so it’s important to understand how that's going to move around. In our business, for our white glove service or our full service, we're doing a lot more fixed price. We can say “Tell me everything you’ve got.” And when they do, we can give them a number, like 5 grand or 20 grand, depending on the size of it.
We just recently put in a bid on a 20-terabyte project and when we put our bid in, they said, “You were a tenth of the cost of the next competitor, and you're offering a better solution quicker. How are you doing that?” And we can say “We're using forensics tools to do something that traditionally people who would do with eDiscovery.” Because we are a full forensics shop and a fully eDiscovery shop and a hosting provider, so we could find the right path to take people down. If I took those 20 terabytes and put it in an eDiscovery tool, it would take forever. But if I take those 20 terabytes and break them into 1 terabyte chunks and put them into ten forensics tools, wow, I can search across that really quickly.
Doug Austin: What else would you like our audience to know about what you and Avansic are working on?
Dr. Gavin Manes: I think we're really trying to deploy this DIY holistic solution, and I don't want to call it a platform because it's a lot of pieces of software, whether it's the old AccessData software or iCONECT out there. We have a lot of SQL in the middle, we use Oracle and we use a lot of technology behind the scenes. We’re thinking about “How can I help my clients seamlessly shift from a forensics investigation into eDiscovery back to forensics and then out to where I need to give an expert deposition.”
Having the right tools or technologies where a client can realize when they’re hitting the “soft fence” where something is telling them they need to reach out for help versus them being totally comfortable. Knowing when in the case that changes and allowing that flexibility, in software, technology, people and price will really help them. The analogy I'll leave you with is a case where we were providing a very large white glove service for them, and the case went into settlement talks for nine months. Five months in, the clients were still paying to have all the users and everything, and we reminded them they hadn’t logged in. They still needed access, but we said, “Let's transition you from full service into self-service. The pricing will be a lot cheaper, and you can turn things on and off when you want.” So, their bill went from a few thousand dollars a month to a few hundred dollars a month quickly. That flexibility is what we're trying to build.
But we can't do is we can't do it by ourselves, we have to have customers come and tell us what they like and don’t like. We want to know because we want to build a product that they want and not just shove it down the pipeline and make them use it. I think that the industry has always been saying, “Here's the only solution.” Now that we've seen the interstate, we can say “What solution do you want?”
Great. Thanks, Gavin, for appearing on the eDiscovery Today Thought Leader Interview series!
So, what do you think? Hope you enjoyed part three with Dr. Gavin Manes! Please share any comments you might have or if you’d like to know more about a particular topic.
Disclosure: Avansic is an Educational Partner and sponsor of eDiscovery Today
Disclaimer: The views represented herein are exclusively the views of the author, and do not necessarily represent the views held by my employer, my partners or my clients. eDiscovery Today is made available solely for educational purposes to provide general information about general eDiscovery principles and not to provide specific legal advice applicable to any particular circumstance. eDiscovery Today should not be used as a substitute for competent legal advice from a lawyer you have retained and who has agreed to represent you.