 from San Francisco. It's theCUBE, covering Conga Connect West 2018. Brought to you by Conga. Hey, welcome back everybody. Jeff Frick here with theCUBE. We're in downtown San Francisco at the Thirsty Bear. We're at Dreamforce. I can't get an official number. I keep asking, but the number they're throwing around is 170,000 people. So if you're coming, do not bring your car. It'll take you four days to get here from 1810. I think the Giants have a home game today too, which just makes things even more interesting. But we're at a special site event. It's the Conga Connect West of it. Here at the Thirsty Bear, three doors down from Moscone South. So we're excited to be here. It's our first time at Salesforce. And to kick things off, we've got Bob DeSantis, the Chief Operating Officer of Conga, and with him, Jason Gabbard, the head of AI Strategies. So gentlemen, welcome. Thank you. Good morning, great to be here with you. So what a cool event. You guys have this thing rented out, I think for three days. You've got entertainment, you've got the disco, I think tomorrow night, some crazy bands. We've got an open bar, food going all day and all night. We actually, we did this last year and we were so crowded that this year we rented the parking lot behind and we built two circus tents. So we actually extend all the way out to the next block. And we have multiple sponsors here helping us to bring their customers and their partners in. So open bar, open food, meeting rooms, demo stations, a place to come and relax and kick back a little bit from the chaos of those 170,000 people just to block away. It's just crazy. So come on down and meet the Conga crew and all the people you have a good time but let's jump into it. Topic at hand is AI. We hear all the buzz about AI, AI, AI, machine learning, artificial intelligence. And what we hear time and time again is no one's just going to go, I just need to go buy some AI really. That's not the way the implementation is going to work. But where we see it in a great example, I like these a lot that people are familiar with is Gmail. Those little tiny automated responses back that email is actually a ton of AI behind those, setting context and voice and this, that and the other. How are you guys leveraging AI in your solutions? You've been at this for a while. AI represents a great new opportunity. Yeah, it really is. Jason, you want to, yeah, sure. I mean, you may not be aware but Conga's actually been developing AI inside of the contract management system for a few years now. And so I came over to Conga in connection with the acquisition of a company I found it focused on AI. And so obviously things are getting a lot more interesting, technology's getting a lot more robust. But I think you made a great analogy to Gmail. And so inside of the Conga CLM, Conga contracts, you'll actually see that we're starting to make suggestions around contracts. So you may load a document in and it's like, you might see a pop-up over in the margin that says, hey, is this a limitation of liability costs? So that's one example of sort of AI working in the background of the CLM. Well, it's going to say, what are some of the things you look for? I had a friend years ago, he had a contract management company and I was like, how, and this was before OCR and it was not good. I'm like, how are you doing this? He goes, no, no, no. If we just tell him where's the document and when does it expire, huge value there. He sold the company, he made a ton of money. Right. But obviously, you know, time has moved along. I've had a lot of different opportunities now. So what are some of the things you do in contract life cycle management? You think of that example as phase one of contract life cycle management. Just get all my contracts into a common repository. Give me some key metadata like, what's the value, who are the counterparties and what's the expiration date. That's huge. And so 10 years ago, 15 years ago, that was the cutting edge of CLM, contract life cycle management. You know, now the evolution has continued. We're in what we think of as sort of the third phase of CLM. So now, how do we actually pull actionable data out of contracts? So having the contract, she mentioned OCR, having machine readable data in a repository is great, but what's actually in the contract? Right. What did we negotiate six months ago that now could have an impact on our business if we knew it, if we could act on it? And so with Conga AI and the machine learning technology that Jason's company developed and that we've now embedded in our CLM products, we can unlock the data that's hidden in documents and make it actionable for our customers. So what are the things that you use to trigger that action? Because the other thing about contracts we always think about, right, you negotiate them, it's a pain in the butt, you sign them, then you put them in the file cabinet, nobody thinks about it again. So in terms of making that more of a living document beyond just simply it's time to renew, what are some of the things that you look for using the AI, what are some of the, are you flagging bad things? Are you looking for good things? Are you seeing deltas? What are you looking for? I'll give you a really concrete example. We recently had a customer that negotiated a payment term to their benefit with one of their suppliers. But that payment term was embedded in the document and their procurement team was paying on net 30 when procurement, sorry, their payables team was paying on net 30 when the negotiators had negotiated net 90. That data was locked in the contract. With Conga AI we can pull that data out, update the system of record, in that case it would have been SAP, and now the payables team can take advantage of those hard fought wins in that contract negotiation. That's just one example. Yeah, so I mean, two obvious use cases we're seeing day in and day out right now. Number one, I'll call an on-ramp to the CLM. So that's likely a new customer relatively new customer at Conga that says, hey, I have 50,000 contracts. I was on the phone this morning with this precise use case. I have 50,000 contracts, really happy to be part of the Conga family, get my CLM up and running, but now I got to get those 50,000 contracts into the system. So how do we do that? Well, there's one way to do that, get a bunch of people together and work for a couple of years and you'll have it done. The other way is to use AI to accelerate some of that. Now, you know, classic misconception is that the AI is going to do all the work. That's just not the case. At Conga we tend to take more of a human computer symbiosis sort of working side by side. And the AI can really do the first pass and you might be able to automate something like 75% of the fields. So you can take your reduced team of people then and get the rest of the information into the system and verified, but we may be able to cut that down from a couple of years to 30, 60 days, something like that. So that's one obvious use case for the technology. And then I think the second is more of a stare and compare exercise. So, you know, historically you would see companies come in and say, if I'm going to sign an NDA, it's got to have, you know, the following 10 features and I'll never accept X, Y, and Z. And so we can sort of key to that with our AI and take the first pass of the document, really do the triage. And so again, while it may not be 100%, we'll get to 80, 90% and say, here are the three or four areas where you need to let your knowledge work its focus. Right. And are there some really, you know, discrete data points that, you know, you call out in a defined field for every single contract because they're always their payment terms I'm at. Obviously dates and signatures, but some of those things that are pretty consistent across the board versus, I would imagine, all kind of the crazy esoteric stuff which is probably the corner cases that people focus too much on relative to the value that you can get across that entire 50,000 contract with a lot of contracts. I mean, I don't know what your view is, but for me, I think it's, you know, follow the money. Everyone always cares about dollars when I'm getting my dollars. And the other is follow, you know, very high risk stuff like indemnities, limitations of liability. Occasionally, you know, you're seeing people interested in changing control. What happens if I sell my company or take on a bunch of finance? Does that trigger anything? Right. What's interesting about contracts is that, you know, there are hundreds or if not, thousands of different potential clauses that could live in a contract. But in general, you know, sort of the 90-10 rule is that it's about 40 clauses that you find in most commercial agreements, most business to business, or even business to consumer commercial agreements. So with Congo Machine Learning, we train based on the sort of use cases that extend that for a specific domain. So for example, we've done a lot of work in commercial real estate, right? So they sort of all, those commercial listed agreements have that core base, but then they have unique attributes that are unique to commercial real estate. And so Congo Machine Learning, as part of the Congo AI suite, can be trained to learn so that we can reduce that cycle time. You know, when we go into our 10th commercial real estate use case, it's going to be a lot more efficient, a lot faster and a lot higher initial hit than when we start training it at the beginning. So, you know, for us, it's about helping customers consume the documents that make sense for their business. And Machine Learning is intuitively about learning. So there is this process that has to take place, but it's amazing how quickly it can learn. You know, you sort of, you use the Google example, I like to think of the Amazon.com suggestion service example, right? Like they literally know what I'm going to buy before I'm going to buy it. That didn't just happen yesterday. They've been learning that from me for the last 20 years or 15 years. And so, you know, we're at the sort of beginning of that phase right now in sort of B2B CLM. But it's amazing how quickly it's moving and how quickly it's having an impact on our customers' businesses. Yeah, because I was just going to ask, where are we on the lifecycle, you know, of the opportunity of using AI in these contracts beyond just the signature date and the renewal date to some of these things. And then also I would imagine you guys can tie some of that back into your document creation process. That's right. So that you again, remove a lot of anomalies and you know, get more of a standardized process. Yeah, so, you know, Kanga provides a full digital automation, digital document automation transformation suite. And that includes, as you mentioned, document generation capabilities, contract management, Kanga AI, and Kanga sign. And you know, so that we're not here yet, but you know, imagine if through Kanga AI, we're able to learn what type of clause structure actually has a higher close rate or a faster cycle time or a higher dollar value for a given book of business. So customer acts is selling their products to consumers or other businesses. And if we can learn, we can, how their contracts streamline and improve their effectiveness, then we can feed that right back into the creation side of their business. So that's just over the horizon. And then the other thing I would imagine is you can get the best practices, both in our department, in our company, and then I don't know if you go, I don't know where the legal limits are in terms of using it anonymized kind of best practice data to publish benchmarks and stuff, which we're seeing more and more, because you know, people want to know the benefits of you seeing so many of these things, you know, what's next. And then do you see triggers? Will someday there, it'll be a trigger mechanism or is it really more of kind of an audit and a just going forward? You know, from my perspective, I think the someday is more, we're extremely focused on the analytics and the kind of discovery of documents right now, but I think sort of looking out over the one year horizon, it's less about triggers and more about more touch points in the workflows. And so really optimizing the contracting process of being able to walk into a company and saying, hey, I know you would like for this to be in all your contracts, but as a matter of practice, it's not. And so maybe we need to abandon that policy and get to a signed document faster. So more of that type of exercise with AI and also integrating with sibling systems and kind of testing what you expect it to happen in the document versus what actually happened. And that may be visa being integration with ERP or something like that. It's pretty amazing because as we know, the stuff learns fast from watching what happened with the chess and the go and everything else. You read some of the books about exponential curves. You know, you'll get down that path probably faster than we think. Well, Bob, Jason, thanks for taking a few minutes. And again, thanks for inviting us to this cool event and everybody come on down. There's lots of people. I'm down to the Thursday there. All right, he's Bob, he's Jason. I'm Jeff, you're watching theCUBE. We're at the Conga Connect West Event at Dreamforce at the Thirsty Bear. Come on down and see us. Thanks for watching.