How Rev Streamlines Legal Discovery With Secure AI (Full Transcript)

[00:00:00] Speaker 1: Sharing my screen now, I'll start with a few slides to let Robert and I introduce ourselves and then we'll transition over to the demo. Just give me a quick thumbs up to make sure you guys can see my screen.

[00:00:10] Speaker 2: Yep.

[00:00:23] Speaker 1: Not sure why I keep muting, it's weird. Just to introduce myself, I work on the sales side of Rev. I help a lot of folks evaluate Rev, make sure it is valuable for the team and then set some next steps from there. Robert, I'll let you intro yourself. You're on mute, Robert.

[00:00:47] Speaker 3: I'm gonna try that again, sorry about that. So hi, my name is Robert Duffner. I'm a solutions engineer with Rev and I help organizations implement Rev and also provide any other technical assistance. On top of that, I am a certified paralegal. I've worked in law firms for over four years so I can definitely give you the perspective of what it's like using Rev as a paralegal. James.

[00:01:14] Speaker 1: Awesome, thank you, Robert. Yeah, at the end of our time, we'll open it up for some Q&A as well. I'm gonna go ahead and start showing the demo environment of Rev. Everybody see this okay? Looks like you can, cool. So I'm not gonna spend too much time on the homepage here. I'm pretty straightforward here. I'm gonna jump right into the meat of it, into the Rev Insights. And so what I'm doing here is I've got all my discovery as if it was just dumped on me over here on the right and I really have no idea what's here. And what a lot of folks tell me is this is kind of where they start. They get all their discovery, they bring it into Rev and then they just wanna understand what is in these hours and hours of audio video, interview files and phone calls. And so what I'm gonna do, where a lot of people like to start, is just to get a brief summary of each file. I'm gonna say briefly summarize each file. So what the AI is doing, since these have already been uploaded into Rev, we already have a transcript of them. So the AI already knows what's in here, but I don't know what's in here yet. So that's why I asked this prompt and what it's doing is it's giving me a brief summary of each file. First thing I wanna call out is these numbers in brackets here. Each of the summary is gonna have this. Reason why it has this is because it's bringing facts that it calls out in the summary straight to that point in the file. So as I hover over this number one, it's gonna take me to that point in the file, same with the number two right here. So it's basically citing its source, letting you know that it's not hallucinating, it's bringing things to you that are exactly from the file. As it's giving me the summary of each right here, say I wanted to just jump into this first one. Say this summary right here that it gave me is really important to my investigative workflow. And say I wanted to drill down into this one first. So I can just click on that. And I'm gonna go ahead and isolate it into its own tab. Now I'm not gonna lose what's going on over here, but I just wanna drill down into this one file first. Now with each file, couple things you're gonna get as soon as you upload them. You're gonna get the transcript and that's what this is right here. Since this is an interview, it's speaker one, speaker two, speaker three, speaker one format. A lot I can do with the transcript. I can go ahead and share it. I can download it. I can do a lot of different things with it. I can also edit it. If for example, I wanna make a correction to say this word right here, this is called the transcript editor and I can make that correction and it'll carry forward to everybody that needs to have access to the transcript. Couple things I'm also gonna get are these AI templates. And you can have several of these locked right here. What this is doing is it's giving me a breakout of what happened in this hour and a half long interview of this officer right here. So right here, I'm at the interview analysis. I really love the way it breaks it out into a general overview, some key points, some notable quotes, more quotes, and then some detailed insights. I'll go back to the transcript and I'll also have this bar right here, which is my AI chat interface. Similar to what I have back over here for the bulk of my discovery. And each file is gonna have its own little chat interface. So I can ask it a question and I'm a little bit familiar with this interview, but I can go ahead and just ask something very detailed. Maybe I don't wanna sit through an hour and a half of this just to find one key point. So I can ask, what did De Leon tell his dispatch? So what it's doing here is it's only looking at this one file. It's not looking at those 20 other files related to this case. It's just looking at this one file. I'll make the output a little bigger here by dragging it over. But right away, I've got an answer here. It says, regarding the pursuit incident, De Leon did not have a gun in his possession. Regarding the pursuit incident, De Leon did not initially broadcast anything about the pursuit to dispatch. He later acknowledged this was a significant failure, stating he didn't broadcast anything at that time. So, and you'll notice that it is citing it in these brackets, which are searchable. So if I wanted to go ahead and...

[00:06:21] Speaker 4: Mystagne's off. I didn't broadcast anything at that time when it was all said and done because I couldn't ID this guy.

[00:06:29] Speaker 1: So what I did there is I clicked to that because I wanted to listen to it myself. And what it did is it sped up the transcript and it took me straight to that point in the transcript, to the audio. If there's video, it would have taken me to the video, but it took me to the audio. And what a lot of folks like to do is if that's something really important for them, is they'll use these pair of scissors here. I know that's kind of small to see. But I'll go ahead and use these pair of scissors just to clip that portion there. Say I wanted to archive this for my workflow. And so what it did is it clipped that paragraph. Maybe I don't need the whole paragraph, so I'll just drag it back a little. Maybe that's all I need right there. Then I click this little save clip right here. So what it did is it took that portion of the audio, say about 20 seconds worth, and it clipped it. And it also clipped that part of the transcript, which I have right here. Now these clips are actually independent from the file. So what that means is I can download just this clip, I can share just this clip, and of course I can rename this clip as well. It's not going to do anything to the transcript. I still have that as my source. But as I'm going through this, I can make these clips that are really important to me if I want to archive those and separate those out from the main source. Other things I can do within the transcript editor is I can make some comments here. If I'm working with somebody and I want somebody to know something, or maybe just some notes for myself. And so I can make a comment on a certain section. Of course I have all the usual power to be able to highlight things. And then of course I can speed up the transcript to 2x, 0.5x, slow it down. I have all that power here and then I can skip within the transcript. Downloading of the transcript, if I wanted to print it out, set it on somebody's desk, this can be done PDF format, txt.doc, all that we got you covered there as well. This is where you would download it and it's going to give me those options here. Say I wanted to keep it in the speaker one, speaker two format, that's how it defaults. We've also got other options here. If I wanted to remove like filler words from the transcript, if I wanted to include the timestamps, I've got all those options as well. And then here I'll choose the file that I want to download it in. Another thing I'll call out within the transcript is you can make speaker edits right here and those will lock. That means is you only need to correct the name once and then the ASR, the automatic speech recognition will recognize their name for the rest of the transcript. So you only need to do that once and then it'll lock that name throughout the transcript. I'm going to come back here. So now my summary is done of each file. So now I have a tight, concise summary of everything in my discovery and I can start to drill down into each individual file. Of course, I can, if there's something valuable within here, I can also come in and within that stack, it's going to fast forward into that point there and then I can hit play.

[00:10:06] Speaker 5: Like I said, chasing him, the guy didn't give a damn, chase.

[00:10:11] Speaker 1: And then of course, I can just clip that little section there. If I need to do that from the whole stack of discovery, you can also do it on individual files or from the stack, completely up to you. The clips will archive in the same place where my discovery lives. So if they live in a folder, the clips will go to that folder. I also have the police reports in PDF format for this case. And I, on purpose, kept those separate from the audio video. So the audio video is here and then my PDF reports are here and all my work is going to live here. You're not going to lose anything, so no worries there. And so what I did with this is you might recall from when I first started the Rev Insights, there's a shortcut called, give me the case chronology. This is not a prompt that I had to type in. It's just a shortcut straight from the homepage. And so I isolated these reports, investigations reports, incident reports, other reports. And I said, give me the chronology on the reports. So what it did is it laid out a very nicely organized chronology. Everything is going to be cited for me. If I wanted to hover over this number six, it's going to take me straight to that point about speeding southbound on the Highway 80-99 Corridor. If I wanted to clip that as well, it can come here. The AI has made it into a text that it can read, which I can copy and archive. And there's my source file here. So this is what I uploaded into Rev. Taking it one step further, a lot of folks also on the investigation side really love to be able to analyze pictures. So what we did just now is we analyzed audio, video, we made clips, we got highlights, we got summaries. And then also on PDFs, we got a case chronology as well. But also there's a power to analyze pictures. So this was actually a very new release. And what I did here is I have 14 pictures in my discovery. I have no idea what's in each of them. So what I did is to get a general idea of what's in each picture, as I said, summarize each image. And what it gave me here was a brief summary of each image. And I can always hover over this just to see exactly what that image is here. So for example, this first one is a mobile, a text conversation. And if I wanted to see a screenshot of that text, there it is right there. So as I hover over, I see a little breakout of whatever that image is right there as well. So analysis of audio, video, PDF, .doc, and images is what Rev can do. I wanted to come back to the templates real quick. So I mentioned shortcuts earlier, for example, case chronology, that's a shortcut that I have locked from the homepage. I can create my own shortcuts here in the AI templates. And this is stuff that we help folks do in onboarding. So I have a lot of different shortcuts here. If I wanted to create my own, I just click new, give it a title, a brief description, and then how to prompt it for that shortcut. This is for if I'm doing a lot of repetitive things when I get my discovery, is every time I get a dump of discovery, I need this, this, and this. Instead of prompting every single time, you can have these templates ready to go so the output is just there for you. That's available here in the Rev Insights, which is our portal for multi-file analysis. It's also available to you here if I wanna analyze just a single file. And you'll notice that I have a lot of little templates right here locked and ready to go. You can have up to five of these. So as soon as I upload a file along with the transcripts, I'm gonna get my AI templates ready to go. Did wanna call out, I forgot to mention earlier, if I really like this template, and this is something that I wanna share with somebody or potentially just print out and set it on somebody's desk, I have all the power to save it to my files, save it as a PDF, print it out, share it with somebody, anything I need to do there. If I wanna edit it, it's gonna open up a little kind of a Word doc interface here. If I wanted to bullet point something, highlight, italicize, I can do all that here within the AI template. Over here as well, if now I have my brief analysis of each file that I've uploaded in the multi-file analysis portal, if I wanted to draft this document so I don't lose this just to share it with somebody, I'll just click that draft document right there, and there's my draft. I've got the same Word doc interface here. If I wanted to make any changes, and downloading this, sharing this, printing it out, is what I can do with this as well. So these are the things that a lot of folks find super useful. It's when they get a stack of discovery, being able to get a quick summary on everything there, and then they start to drill into those files, take it a step further within those files, creating and archiving separate clips that are independent of the files, and then being able to look at images as well and get a breakout of what those images are, specifically if they're looking at pages and pages of text messages, maybe they're just looking for one name within that text message. Our AI does that extremely well. At this point, I think we can transition over to some Q&A, and I'll leave this, I'll keep sharing here in case somebody has a question about one of the AI outputs here.

[00:17:10] Speaker 6: Hey, James, I do have a question here.

[00:17:13] Speaker 1: Sure.

[00:17:14] Speaker 6: The question is, if I need a second person to review something on a sensitive case later, can I grant access to just that one case without exposing anything else on my account?

[00:17:24] Speaker 1: Yeah, of course. So back to the homepage here, I mentioned I would come back to this. The way to do that is you would create these workspaces here, and right here is where you can add members to that specific workspace. So there's only one person in here right now. If I wanted to add a new member, just click Add Member, add their email, and then give them the type of access that you want them to have. They're not going to see my files here. These are for me. They're not going to see the shared with me, and they're not going to see any other workspace that I haven't added them to. So they'll only be able to see the workspace that I'm adding them to. Great question.

[00:18:12] Speaker 6: Okay, I've got another one here. How does this compare to just using a general AI tool and being careful about what I upload? What's structurally different about Rev's setup?

[00:18:24] Speaker 1: First call out, I would say, is our security. So it is built for legal. It is a closed-loop system. So what that means is we don't have access to the data in here, even as employees of Rev. You own your data completely. If you delete something, it's gone forever. We have no way to retrieve it. The next thing is, is our AI, the way AI gets better is it trains off of the data, and it starts to familiarize itself with patterns and get better and better. What Rev does differently for legal and investigations is our ASR is not training off your data. That's why we have different levels of compliance in terms of CJIS, HIPAA, and the way we encrypt your data. So it's closed off to you. You own the data. We're not training off of it. We're not looking at it. It's completely yours.

[00:19:23] Speaker 6: Okay, great. That's everything that I have in the Q&A. Again, attendees, if there's anything that you wanna ask, just go ahead and ask that now while we're here in the Q&A portion. If you would like to follow up, we'll be sharing James' contact information as well if you'd rather send those questions over email. Oh, I see one more here. What are the differences between tiers that Rev offers?

[00:19:57] Speaker 1: Great question. It's really gonna come down to security and the type of onboarding that you'd like. So if you're on our pricing page, the option on the far right, the custom plan, that's gonna come with those security protocols that I referred to earlier. If you need CJIS, HIPAA, if you need a closed loop system where the ASR is not training, or looking at your data, that would be that option. That option also comes with an account manager on our end that's gonna onboard and train your team and create your workspaces and give you some guidance on what we see working really well for customers just like you guys. The other plans, although they're great, they're more intended for individual users. The security isn't quite as tight, but they're gonna be there for you. The ASR isn't quite as tight. It's not CJIS or HIPAA compliant. You don't have a dedicated account manager and the ASR is training off of your data by default.

[00:20:56] Speaker 6: Okay, awesome. Thank you, James. I think that's everything we have from the Q&A. Perfect.

[00:21:11] Speaker 1: I'll stop sharing here. Maybe it's a good time to open it up for Robert. Robert is actually somebody who has helped folks make digital transformations. He's consulted attorneys, a lot of different types of investigators, and he's got an interesting perspective on that end, and he does work here at Rev. So we're really glad to have him. Maybe he can just share a little bit of what he's seen working.

[00:21:38] Speaker 3: Yeah, James, that's a good point. A couple of things resonated with me that you covered in the demo from an investigation perspective, which is, I thought you put it well when you said that initial dump of files that you get, and then just being able to get a sense for what am I dealing with before I start digging down into every single file, right? And just having that overview is really, really cool. I kind of like this. I like the idea of a case chronology. I also like that it included a file manifest. What are all the files that I'm dealing with? Because one of the challenge, the technology is great, it's easy to use, but it's really how organized are you with your files, right? So that has nothing to do with the technology. That's really about how you execute your workflows inside your firms, so.

[00:22:31] Speaker 1: Yeah. Great call out.

[00:22:36] Speaker 3: And then I think the other thing, kind of to the other question, they said, why use Rev versus a frontier model like just OpenAI or one of those others? And I thought you really nailed that point pretty well. The feedback that I've gotten from law firms is that they like that Rev keeps everything on the rails, right, that it doesn't mean that those LLMs don't have a place in things that you may need, but when it comes to doing these types of legal workflows, having everything contained, secured in a single environment, and not have to worry about everything that an LLM is trained on showing up as a hallucination. So yeah, those are, so you can actually coexist with those in an environment, it's not an either or, right?

[00:23:32] Speaker 1: Yeah, great call out. Thanks, Robert. Well, Olivia, if there's no other questions.

[00:23:42] Speaker 6: Yeah, I think we're good to wrap up, unless there's anything else you guys wanna cover.

[00:23:48] Speaker 3: No, I think that's good.

[00:23:50] Speaker 6: Okay, cool.

[00:23:50] Speaker 3: Thank you, Olivia.

[00:23:51] Speaker 6: Yeah, well, thank you everyone for joining. Like I mentioned, we'll be sending out a recording of this following the session, and if you have any other questions, please feel free to reach out.

[00:24:02] Speaker 1: Perfect. Thanks, Olivia.