Name: AI and Contracts: Shifting Insight Beyond Legal
Uploaded: 2026-03-18T18:16:06.934Z
Duration: 59 min 8 s
Description: AI and Contracts: Shifting Insight Beyond Legal

Transcript for "AI and Contracts: Shifting Insight Beyond Legal": Linda, and Navin, I think is joining us right now on stage. Welcome to the webinar and thanks for joining us. Good morning. Thanks for having us. Yeah, thank you Phil. Hello everyone. Now, the first question that I always ask, and I'm gonna ask this, I'm gonna pose it to all three of you actually. Linda, I'm gonna start by putting you on pots to eat first of all. The question is, did you find procurement or did procurement find you? That's a great question. As with most of us procurement practitioners, I think procurement found us for those of us sort of started out earlier on maybe two decades ago. I actually started out as a public accountant auditor and so was tapped on the shoulder by the co founder and president of a small subsidiary that GE Capital owned at the time called Genstar Container. And the founder basically said to me, hey Linda, we need you to start up a sourcing and procurement organization. And I said, I've never done this before. Why me? And he said, why not? Then he added, don't worry, we won't fire you. And so that's how I got there. We got into procurement. You had a bit of a safety net there. And I guess the rest is history. The rest is history. We've been doing this as a procurement practitioner. Yeah. For over two decades now. Yeah. How about for you? How did you find your way into procurement? So procurement taught me through a series of roles where I realized love the intersection of business strategy, supplier partnerships and operational excellence. Over time, gravitated toward roles where procurement could be a strategic value driver, in addition to not only being a cost reduction function, but actually being a partner and a consultant to the business. And last but by no means, we, how about, how did you find your way into this profession? Know, Phil, I think in my case, procurement found me usually to yell at me review the contract after. So I am a recovering attorney myself, and so, you know, I kind of classic story, I moved into legal ops over the years and as my career developed there, I found myself working closer and closer with procurement teams as well for obvious reasons that we're going to talk about today. So, yes. So every time that stakeholders shouted at us because we didn't have a contract signed, it was really fault. Pretty much, pretty much. I was that guy that was taking too long to review something. Right. You know, needed to sign XYZ with our suppliers and Navin's the bottleneck. So apologies to all my previous colleagues. Hal, I want to ask you, you work for a tech company, does working for a top tech company actually impact your team's mindset towards digital procurements and digitization of procurements? Sure, yeah. So at Rocket Software being that we're a software driven organization, our teams expect automation, data transparency and speed. So there's a mindset, if we can digitize and streamline something, why wouldn't we? That expectation puts the procurement to stay ahead. So constantly evaluating tools to integrate, automate and unlock insights, especially in areas like contracting. Yeah. Yeah. There's no excuse for, you know, kind of traditional methods when the rest of the organization is moving full, full force ahead with software. Correct. Linda, your title is really interesting. I remember it was probably a couple of years ago or maybe time flies and it was even further back than that when you were given your current title, which is Chief Value Officer. I think that it suggests something important about the attitude towards procurement at Box. And I wonder if you could just talk a little bit about that. Absolutely. I think it took a global pandemic for all companies to understand and value, what procurement brings to the table and what procurement can do for the company. To Hal's point, if we're like the internal consultants to the company, we bring more than just contract negotiations and savings. And so as part of that, with Box moving to AI first, all the AI craze that's out there in the market in the last couple of years or so, we sort of took a refreshed and reimagined look at what procurement can be at Box and one of them was absolutely we contribute more than just the dollar savings from the contracts. We really also look at process improvements if it's bringing in efficiencies and productivity and so in order to sort of encompass all of the attributes that procurement can add and contribute to the company's value and everything. So when I got the opportunity to name myself, rather than I was already Chief Executive Officer. And co founder of the company and also the CFO basically said, well, you know, you can sort of if you could rename your department and what you do, what would you how would you name it sort of in the vein of this whole reimagining with AI first and everything. And the best I could come up with was Chief Value Officer. And so here I am now, I'm now responsible to have to also report out metrics around efficiency improvements and productivity improvements as well. Do you find that that renaming changed perceptions? I mean, obviously you've got to do the doing to really change perceptions, but did it, kind of put the idea of procurement in a different light about, how you drive value for the business? Oh, absolutely. Just by the title. So now my team members are not procurement or purchasing specialists or even sourcing specialists and sourcing managers. We are value partners to the business. And so my team members are value directors, value partners, just that name change. I think for a lot of us who have been doing procurement a long time, we complain about how we're always sort of the second class citizens, the stepchildren, and we don't always have a seat at the table. Well, now we have a seat at the table. And so if we didn't like what sort of purchasing and procurement kind of meant and the reputation of the nomenclature that came with that wording, definitely the having words matter, definitely have value, right, as part of your title and part of your name, you reach out to either internal business partners and or your external provider supplier business partners, definitely has a different expectation and different ring to it. Yeah, we could probably spend the next twenty minutes on naming, but I'm such a big proponent of branding and branding of the procurement organization. And yes, we have to then it's all well and good rebranding and having a different name. And then everybody, the skeptical stakeholders are going to be like, okay, well, changes? So you have to follow through with that, but it's on us to drive that change of perception. It's not to wait for somebody else to decide one day that actually procurement is valuable. You know, let's treat them differently. Right. I think sometimes we just wait for that. Agree. I fully agree. So it's up to us to kind of change that perception, but then I also say we earn our credibility and reputation one deal at a time or one project at a time. It was the old fashioned way. Navin Mahavijiyan, to what extent are you and your team as you love talking and working with procurement as opposed to legal and finance folks? Do you see the procurement is more and more involved in thinking about digitization when it comes to the contract kind of management stack? Actually, know, Hal, great question and before I say that, you know, Linda, you've inspired me. I think I want to ask Agiloft about a title change to a cheap cocky person maybe, but you know, actually answer you Hal, and it goes back to something Linda said as well, you know, I've seen kind of procurement teams move from being the neglected department when it comes to CLM, which is largely driven by legal, to kind of being one of our core, sometimes the main stakeholders. So at Agiloft, we work very much so with procurement teams and CFO organizations, as much as we do legal and sales teams too. So it's definitely trended in a way that procurement is heavily involved in contract data and CLM implementation, management and selection as well these days. Do you find that decisions are made, today more holistically across both the sales and the procurement continuum, as opposed to, you know, we're going look at this from a sales perspective and we're going to look at this from a procurement perspective differently? No, I think definitely differently. What I see is that procurement organisations today are a lot more mature when it comes to managing contract data and leveraging it. And so, tends to be a lot easier to work with procurement teams, whereas sales teams, they have a different set of goals. And of course, it's a completely different piece from procurement. Sad to say, most sales organisations aren't as mature when it comes to that, with exceptions of course, like everything else, but a lot of them are focused on velocity and that muscle around data management and developing and utilising structured data just isn't there yet. Can we take a quote of that? Sales is not as mature as procurement. I was gonna say, we always knew it. Well, money on all procurement's conference rooms around the country. It's going to make the rounds on social media. Navin says salespeople are immature. Can't imagine why people have been yelling at you Navin. That one last question I do want to ask Navin is we talked about AI when I introduced the topic. Talked about AI all the time as we know as a society right now, when we're talking about AI today, just the context of the conversation, are we talking more generally about AI or are we talking about generative or arrogantic AI specifically? I hate those definitions a little bit sometimes, Phil, but I think we're going to talk about both within the context of procurement, but also I would say it doesn't matter as much. When I think about AI, I think about the problem that we are solving, not what we call it specifically, and so I think definitely that will be part of our focus as well. Okay. Kelly, I'm going to hand over to you in a sec, but I just want to remind everyone We who's joined do actually have a download that's posted in the chat window. It's called Unlocking Contract Data with CLM and AI from Locked Files to Living Intelligence. And please, if this is a topic of interest and obviously you have joined us on the webinar, so hopefully it is, please check that out. There's a lot more insight in that report as we go through the conversation. Kelly? Sure. And just to give people added incentive, if you download it, it's actually just a big printable poster that says procurement is more mature than sales when it So comes to everybody's going to want a copy of that. And share that with your favorite sales reps. This Yes, sending her to an East side, West side type I of like it. I like it. When we start to think about taking everything that we've heard about AI and applying it in the procurement tech stack, Linda, you spoke about how at Box, there's been this open embrace of the potential of AI. Are there any specific places, layers, spots in the procurement tech stack where you prioritized experimenting with what AI could do? Absolutely. And Navin, to your point of being sort of starting out in legal ops, at Box, we have, between procurement sourcing and legal, we have a hand in glove, one of the best, if not the best working professional relationships I've ever had in my career, where legal departments usually care a lot more about the legal terms and conditions and procurement sourcing value. Or we're more concerned about the commercial terms. For example, there may be limitation of liability terms that we care about from a legal perspective, but for procurement and sourcing side, care about the termination rates. At the end of the termination, is there penalties and so on, these sort of the commercial impacts to the company. And so that's why together, in partnership, we're hand in glove at Box and we work very well together. In that area, with the development and the current maturity areas of where AI and agentic AI is going, definitely, like we call it CLM, but CLM is very broad to me. There's the whole red lining and terms for library before you sign. And then after you sign, which is where I live and breathe every day, after you sign, there's the ability to be able to extract intelligence and actionable data from that contract. And so that's where I prioritize, if you will, what we can do in my organization, where we're expected to do a lot more with a lot less and a lot faster. And so we spend a lot of our times and hopefully, Navin, you would agree, in terms of like comparing data sets, comparing contracts, comparing renewal quotes and proposals to what we had agreed to, or even relooking at our master agreements that have been around or in existence and signed five years ago, ten years ago, nobody really bothered to relook at it. And so from a procurement perspective, that lens and that ability to be able to extract what changed and what are some of the terms we really care about was number one for me. And then on my legal team's partnership side, what they care about is also that what are some of the more risky terms that we agreed to, right, as a company? And what were some of the terms that we had agreed to more often so that we can develop our own clause library without really having to implement an actual CLM solution? Now it's controversial, people might disagree with me out there, but I believe like if you have the ability to be able to do that extraction and do that comparison internally, you really don't need that CLM solution on the front end to help you with the red lines and the clause libraries, because you will have it by extracting it out of what you've already done in the past historically. And so that's what we've worked with at Box. And I have to say we do have a very, very extensive library of what was agreed to, what were some of the more risky terms, and being able identify where those risks are and go, hey, we need to renegotiate those terms when the renewal comes up, or sooner, and being able to manage those risks as well as commercial terms more proactively versus waiting and being reactive. It's such an interesting use case, not just because it certainly applies, but looking even beyond, we're not just going to say we're using AI in our CLM, looking at specifically a part of the process, it also seems like it aligns very well with your value orientation, but you're not just looking to tactically complete something faster or even more efficient. You're thinking about, okay, big picture across agreements. How does this affect the business? How can we start driving more value your point? Absolutely. And the timing of it too. So most people, most organizations wait until Oh, our contract's expiring a week from now, two weeks from now. And usually in my past experiences, you know, I get calls from the stakeholders and our business partners, come Linda Chuan and your team, come and help us negotiate something that was expired a month ago. So how much leverage do you think you have then, right? And so with the AI and generative AI capabilities now, we're actually moving what I call up and to the left of the value chain, the value process, where we are partnering with our business partners in way advance of even just the notification period. So people kind of focus on the expiration date of the contract. But no, there's a notification period that you can either notify you're going to continue with the supplier partner or you're going to notify that you're going to terminate. And so you have to get way ahead of that. So we have found it took me, I've been at Box about six and a half years now, but it took me a few years to even sort of cross that journey, cross that bridge to become more of the advisor, what Hal had said earlier in terms of advisor and consultant to the business. Now Hal, when you think about your CLM, and I'm sure the decision making and evaluation process that you and your team went through, were there specific either opportunities you were looking to seize, or problems or challenges you were looking to solve for that really motivated the group's decision? Yeah, we wanted to get past the contracts of static PDF, that they're just sitting somewhere. Contracts were in people's email, they were in shared drives, SharePoint, you name it, they were everywhere. And there was no unified intelligence layer looking at all of this. So our objectives were we want to centralize all of our agreements, wanted to improve our visibility, right? So we want to know what our obligations were, what renewals look like, what was our risk? What clauses were in each contract? How often were those clauses in the contract? If we wanted to terminate something, do we have termination for convenience? Do we have to terminate for cause? So, understanding how all of those contracts were working and then workflows. So repeatable processes that we could go through based on preset parameters, level of risk to the organization. Is there data being held outside the organization? And then faster cycle times. I mean, always says legal is taking too long, right? Weeks, months, whatever the case may be. So trying to bring down those cycle times to get contracts and things out the door. So you don't miss opportunities, A discount or something that's pending. Suppliers want to say, if you get it done by this date, it'll be this. If it's after that date, it'll be that. So, making sure that we're trying to run on time and try to close things as quickly as possible because the business needs to get things moving. Yeah. Now, was it a process predominantly driven by procurement or did you also have cross functional stakeholder groups that were also evaluating or giving feedback or prioritizing objectives? To what extent were the requirements and objectives for the solution company wide? Great question. So, we brought in stakeholders from, of course, legal and procurement were kind of like the co owners of the project. And then we brought in folks from sales and other parts of the organization, folks that were contract heavy, We would get a lot of requests from, we brought those folks into the fold. Trying to frame it as a shared system of record across the organization. At the end, care gets better commercial insights and legal gets better control and version consistency. Are you currently either using or testing any form of AI within your CLM? We are testing some of the Anthropic and some of those other AIs to see what they generate, put them on top of the data and see what comes out. So we're doing some of that experimenting there. We're also experimenting with some of the other tools that are out there that are more, not necessarily CLM, but more legally driven to kind of see what kind of things they bring out. I think the other area where we tend to struggle is when you're doing an M and A, right? And you're absorbing all of these documents from the target company and trying to sort through all those. So being able to bring all those in and surface what are the challenges or what are the business arrangements or is there any overlaps? We already have an agreement on our side that's on their side. Solving some of those things, trying using AI to surface those types of data points. Navin, when you think about the conversations that you're having across companies, across procurement teams, Do you tend to hear from people like, well, they told us we have to put AI somewhere, so maybe we'll try it here. Are you hearing about very specific things like some of the objectives that Hal and Linda spoke to? I don't know if that was me that froze or Kelly. No, I think Kelly froze. I can take a stab at asking Kelly while she gets back online. Think it's really use cases, like the most common Exactly. Use You know, sad to say sometimes it really winds up being a let's slap some AI on and hope for the best. But what I found to be the most effective use cases are, well, surprise, surprise, data extraction, right? And while that sounds boring initially, I think where AI is different is you can do data extraction at scale and you can pivot really quickly. In fact, we have a customer event later this week where I get to kind of put on my pretend stakeholder hat and throw a fire drill request. We need to figure out this data from our entire repository around data breach and stuff like that. And if you haven't been tracking it already, that's a huge lift without AI in the mix. So I think data extraction and going back to, I know Linda and Hal, you both touched on this as well, you manufacture that single source of truth far more effectively with AI. Can't help you with the guy that keeps all the signed contracts in his drawer, but once you get them into that folder, your big repository, it's really, really powerful. But also for risk management, you know, one thing with AI being kind of like sitting atop a well structured repository, you can analyse risk at both a scale and speed that you never have been before. It can be something as simple as which contracts can be terminated by the vendor for no cause, just for convenience, that's risk. I'm giving you a really simple example because otherwise we'll be here all day as well. The other piece is playbooks and Linda, I know you touched on this. I think of this use case as there's two flavors here. One is we can build our playbooks based on what we actually do, what we actually negotiate using all of that structured data that we've been developing, maybe with AI as well and what that helps you do is that you move from this is a playbook built by legal based on what they think are our preferences to this is a playbook built by all of us together based on what we actually wind up wanting to do with our vendors and suppliers and so that's really interesting. And of course, the other piece of it is the red lining using playbooks, which I know lawyers don't like as much, but one really powerful way you can use that in procurement, and this is a great AI use case as well, is letting procurement folks pre negotiate the contract or negotiate it fully as well, you know, because the skill barrier to redlining these contracts is much lower and you can still maintain your standards as well. And then finally, I think the most common, I think favorite piece of it is some form of automation of the intake and triage layer for contracts, whether that's as simple as having AI fill out that gigantic DMV form we sometimes build as our contract intake, know, wink wink, I hate doing it, but sometimes we need to, right? And also just analyzing the contract initially to determine what it contains, what it doesn't, and creating a summary or just pushing it back. And so that's kind of, I think, Phil, where we get into that differentiation between agentic and generative AI as well. But those are, I think, the most popular use cases I see with Legal and in my opinion, also the most effective for procurement teams. When it comes to those playbooks, do you find that, you know, we can now use AI to self improve, if you will, or to kind of, to look at what was actually used and then analyze, well, this is the real world, you know, versus what playbook may say. And either, you know, it's kind of self improved kind of feedback or at least tracking, this is where the playbook worked and this is where the playbook came up short. And therefore it gives us insights to then go and continually improve, perhaps more manually, which is I'm sure what the legal folks are a little bit more, risk averse and they wouldn't want something self improving, But do you find that we now have that ability to start doing things like that as well? Oh yes, yes. Like at higher maturity levels, this is obviously not the very first thing you do when you're implementing AI in a CLM system, but you can look at what we've been negotiating, where do contracts tend to stall, what are we negotiating or what are we redlining in these contracts, and you can then make, I'm going to sound super corporate, but actual data driven decisions around your playbooks as well, where do we really need to push back on this if we keep making an exception 50% of the time? And maybe the answer is a lower level of approval, different ways to skin that cat, so to speak, but using AI tools to identify where the gaps are and where you can negotiate more effectively is incredibly powerful as well. Yeah, thanks Navin. Hal, I want to come back to you and talk a little bit about, as you think about AI enabled CLM in the context of business outcomes, which both you and Linda have talked about being so important. It's not just about saving money. It's what are we enabling for our business? How would you actually start to think about measuring impact? So, we're shifting from kind of, let me go find that PDF to I have real time surgical intelligence that I can act on. Renewal terms are on a surface with alerts. So we have plenty of runway to work on those and negotiate those pricing as those identify proactively. So we know from a budgeting standpoint, what's coming down the road in terms of price changes. We can track buying commitments, make them visible. If there are any rebates or things attached to those to make sure that we're collecting on those. And then we can understand historically what our obligations were, and we can use those to create our sourcing strategies and then come up with go to market scenarios. Now, as you kind of gather this information, gather the insights that you have within the tool, Is this procurement only or do you give access to that to others from outside of procurement? So we do make the information available to others outside of procurement as well. Business owners are tagged on contracts so they can look at their contracts and also get that information. Alerting is done as well just outside of procurement and legal, but the business owner as well as alerted when his or her new renewal is coming up. So they kind of have some headway as well to prepare. And then it also gives us an opportunity to start a conversation about, are we continuing down this path? Are there opportunities we should be looking at? Depending on how strategic the country is organization, you could start this process you know, a year out in some cases. Do you find that that's reduced the number of last minute fire drills that you had of people coming saying, Hey, I just realized that my contract lapsed two weeks ago. As Linda mentioned earlier, help me. Do you have less of those now that you have a little bit more, the stakeholders have more insight into where things stand? Yeah, so there's a pipeline in the calendar that's created. So everyone has full visibility in terms of what's coming up and when. So yeah, the O's or the Agnesys one, those have definitely been reduced considerably, but more the accepted than the norm. That's good to hear because that was always the bane of my life for sure as well. Linda, what kind of lessons have you and your team learned as you've started to put in place AI enabled solutions, you know, around CLM? I am really lucky, to be honest with you, to be working at a company where we provide this content platform, and we're agnostic to all these AI tools. I get to play with the clods of the world, the Anthropics and OpenAI and Gemini and all of that together. But I do also failed very quickly on a lot of things as well as we're going to this AI first. And what I really learned that I want to impart on everyone is that usually people take a very sort of a pain point problem niche approach at applying AI, whether it's UCLM or any other parts along the process. But AI really exposes those problem areas, those pain points even more. Really hyper transparently shows that the problems and the gaps between the processes and the systems and so on that you may have in your environment. So what I'm learning is we're taking a it takes a little bit more time, but to enable to reimagine the whole process end to end, you really have to go and go through a design phase and enable AI where it's more relevant and applicable, where AI could be helpful. So you can't really just throw AI at the wall and see what sticks. And that's been sort of the approach. Gartner, Forrester will tell you there's a high percentage over 85, 90% of our most people's AI projects fail and it's because of that. But if you intentionally redesign, reimagine your process the way you'd like it to be and pick those areas where you can enable AI to come in and do some of that automation for you just do some of that analysis for you, then longer term, your ROI is going to at least be over 30 something percent. I'm experiencing over fifty, forty, 50% of productivity and efficiency gains. And so it's just like, sometimes you have to go slower, a little bit slower to go fast later. And so it's good old fashioned like process reengineering, but to now then add on enablement for AI. Do you find that, do you look at redesigning processes? It's an age old question. Know, redesigning process around technology versus, you know, how do you embed technology into your current process? Have you taken a specific starting point with that to say, okay, let's kind of think about process reengineering first, based on what AI can do, or this is our process, which bits can we make more effective? Well, depending on the company, of course, I'm taking a very much of a sort of Six Sigma white, what do you call screen, label, white A whiteboard or something. Kind of starting from scratch. How should that process work? And so I've been speaking about this a while now in terms of we in procurement, maybe even legal as well, have been voluntold, I guess, or been sort of, yeah, basically told by providers and software providers out there like this is how you should work. This is how you should do your job. But with this whole AI capabilities, we are now for once, first for first time in my career anyway, like this was where I get really nerd out nerding out and really excited about the practice is that now we have solutions that are that are flexible enough to apply. You can apply it to how you want to work, your business process should be. And so that's how I've been designing sort of reimagining your process. How should it be? And what are all the things that you need along the way from a automation functionality perspective? Then layer on what you have in terms of capabilities. Is it some of your especially at VOXX, what our internal capabilities are, right? Drinking our champagne first, and then add on, well then if we don't have those capabilities internally, because, you know, because we're not going to be VOXX isn't going to be purpose built for every use case, But then what other third party AI technologies can we then partner with? And so I'm currently in the middle of a big, huge enterprise wide transformation of source to pay, sourcing to pay, to end process where, to Hal Wechter's point, I probably have, and it's the most complex, we're very collaborative at VOXX, but it also brings in a lot of collaborators. I have over seven different departments who wants to have a say as to whether we bring on a supplier or not. And so then how do we, I have so many masters and how do we make sure that we keep each one of those masters, you know, satisfied to a certain extent. And so it's like having gone through this exercise, this is probably my fourth or fifth, re sort of designing and reimagining, re implementing end to end solutions. This one has got to be the most rewarding because of all of its AI attributes. That's new. You're kind of building something for the first time. Oh my gosh. Yes. Well, then you have to think differently. All of traditional ways and again, I'm going to be controversial again. The way we've been taught in terms of like, you know, spend management, category management. Well, believe, I strongly believe that's because that's how the old Oracles and the SAPs and even Coupa, the old solutions, that's how they were architected by engineers who really don't do our job. And so now with the capability and the flexibility of AI and technology where we can go into unstructured data as well as structured data, I'm actually reorganizing or redesigning my department and trying to figure out how do I then with if I can't add headcount, right, if I have a limited resources and I have to do things faster and quicker and so on, what is the best org design to enable AI as well, right? To be able to do more faster with less, even lesser, even less. It's a lot, but it's also very exciting. Really does challenge you to think differently. I mean, AI just, yeah, it's like, it's not the same old traditional way anymore. You know, that's what I love about this moment more than anything else as well, because, you know, we operate in this or we have operated in this box where we believe that, you know, this is what a process looks like and it's beginning to end. And how do we optimize that process? It is now about what additional go back to your role, like what additional value can we drive for our organizations involve our supply partners? And that's a completely different discussion, but it makes you think completely different about how do I architect what I'm doing? What technology do I need? What's the skills of the people that I need? And, you know, I understand there's a lot of fear in that because it means there's a lot of unknowns, but for me, there's a lot of excitement that, you know, we've kind of over the last twenty five years wished for this moment and the mentored that we couldn't do these things. So now is the time where we actually have the ability to go and do something different. Absolutely. Now it's here. The time is here. Now we're at the table. Now we've got to deliver. And so how do we do that? Somebody has said, the fear comes from people not knowing and the fear comes from people thinking AI may take your place and take the place of a humanoid carbon unit job. But I believe it's not that they'll take over, not yet, but it's definitely a supplement or complementary function where they can do the work or AI can do the technology can do the work much, much faster. What used to take months and weeks of hours can only can take what thirty seconds now. But you still need that human factor to, you know, for the judgment for making sure that AI isn't hallucinating, but you definitely need that complimentary functionality for both. Navin, I want to bring you back because you see so many different companies kind of across the maturity spectrum, would say, as they're looking at one, you know, implementing, but also, you know, working with CRM technology on a day to day basis. What kind of things do you observe from, you know, those that have been able to take go from like simple contract retrieval to actually an expanded focus on insights. And what kind of needs to be present for an organization to make that leap to really go to the next level of what they can do with their contract data? Well, I'll give you the straight up answer right away Phil, which is you need infrastructure. And then now I'll explain it, right, because definitely there's several thresholds you have to cross and the first is always let's find the contract. Can we find contracts that we are looking for? Because with less mature organizations, it really just becomes let's get the contract signed, file it away somewhere, and then worry about it when something goes wrong, or when we realize we need to renew it, or that it auto renewed without us planning to auto renew as well. So that's really that first threshold that these organisations have to cross before they kind of become more advanced. So building that single source of truth so that you can just find the contract and the more advanced level of that is that you can find answers. Because when you're looking for a contractual document, you're not really looking for the document. What you're looking for is an answer. Can I assign this? Can I cancel this right away because I need to change suppliers? Stuff like that. And so that's the next level and of course getting to that also requires a bit more structured data or rather if you want to get the best answers you possibly can. You need that structured data that you can develop through your workflow because you are collecting information that isn't in the body of the contract, the contract document itself and potentially integrating with other systems as well to pull data from there. So, you know, said data probably five times too many, but I think that is where it is. So when I look at the kind of teams that are doing it really well, it was less about willingness or kind of being willing to change or adapt, but rather the ability to both build and have that infrastructure in place so that you can kind of perform at that high level of risk management, being a proactive procurement team. And of course, you know, using the technology to solve problems versus just having it there and having, you know, a format you follow, but at the end of the day, not that much different from an immature organization, so to speak. So I'll use the word data a couple more times. That was one of them. The next one is, how do you connect workflows with good data? Like how do workflows drive good data? Because the data is a, it's been an age old problem for all of us, not just in procurement, I'm sure beyond procurement. Yeah, no, I think a lot of folks have a love hate relationship with workflows, right? But I think of them as these are the train tracks, you know, you're using them to guide your process, they are your process, but also they provide a way for you to effectively collect data in a structured way. I said data again twice now, but you guys get what I mean. It really just provides a format that the rest of the organisation can use and it isn't all ad hoc, all over the place, and so out of that workflow you can have governance around what you build and how you build your infrastructure. So at its core level from that, the data discussion, it allows you to generate structured data more effectively. It does a lot of other things, obviously, from an efficiency standpoint, from auditability and things like that. But when I think about contracts specifically, I know that what I have in the contract is only part of the puzzle and it's great to get all that data, but there's a lot more that I would want as a legal person after the fact, with the different operational departments as well. So that's the power of a workflow to make sure that happens. And the workflow can help normalize the data and make sure that you're capturing certain data elements so then it's readable and kind of consistent on the output side of the workflow. Exactly. And it also keeps people from emailing me asking for stuff. They can just send the it intake form and stop calling me. Know, as introverts in procurement, as fewer phone calls we can get the better. Hal Wechter, I want to come and talk just a little bit about your journey at Rocket Software. And, when we talk about making contract data more available, more accessible, more actionable, what does that actually mean in practice for you and your team? So, through the workflows, we gather the information and we have all the stakeholders involved. So security is getting information they need right from the supplier, SOC two or the certification, things like that. So they get to do their process. Legal see any information they need so they can redline the contract and make those changes. And procurement is getting the information that they need from a commercial standpoint. So I think when you build the strong workflows and you have the right data gathering elements that helps streamline the process and really gives everybody back time. So it's a time saver at the end. Now, as you went through your implementation, what were some of the choices that you looked when you looked back that kind of had a material impact on the success of your rollout? So I think we started with use cases, right? We didn't look at features or functionality. We just said like, what is the business problem we're trying to solve for? What are the pain points that we're trying to reduce? We didn't over engineer the model from day one, right? We built it simple and then build it out from there. Make legal a partner, not an approver. So make sure legal is part of the process, not just someone is checking a box. If it's not user friendly, it's not going to be adopted. So people want to use things that are easy to use. Keep an eye on the workflows. That's real efficiency gains come from. So if you see there's bottlenecks or the workflows aren't moving the way it should be, go back and redo them. The systems today are a lot easier to reconfigure. It's a lot of drag and drop. There's a lot of programming or development that hasn't done anymore. So you can change things quickly on the fly. You don't need perfect data. Start with what you have. AI can actually help you enhance that data and improve the quality over time. AI doesn't replace judgment though. It can help admins. It can help accelerate things. But you still need the human focus on strategy, risk and negotiation. And again, CLM is just not for legal to share business system across the enterprise. Do you find that it has strengthened your relationship with legal by implementing? I have. We're getting faster cycle times, request a signature, getting fewer surprise renewals, getting higher compliance with preferred terms. And we're having better conversations in terms of not only legalese, but risk of the organization, how to manage that. And I think that's so important because that's where we elevate the role is by the different level of conversations that we brought into and how we make some of the easiest things as easy or as difficult as possible for our stakeholders. And the easy that we make them, you talk then about, the ease of use of the tool that just, that makes procurement not look at, not be looked at as a blocker. And it'll be looked at an enabler, which then allows you as well then to get involved in a lot more spend categories because so many times they think about Customers, they come to you and said you haven't come to them. Linda, I saw when Hal was talking about data and not necessarily having perfect data, I saw you nodding furiously. What's your kind of philosophy on that? Oh my gosh. I think Hal, I agree a 100%, thousand percent with Hal. I think you hit the nail on the head in terms of you don't need perfect data. I don't think we'll ever solve the whole quality of data issue, having to normalize and dedup and all of that stuff. Just start, I believe you can start using AI and just start with what you have and you can, you know, implement and as you the longer, the more you use the information, the data and you cleanse it along the way, the data will get better. And I think with the full functionality of AI, exactly, you don't need perfect data now. But I do also believe people, we, should take a little bit of make a little effort in terms of if we want better data, that there should be some kind of a data validation on the front end, again, Six Sigma thinking of dirty data in, dirty data out. And so if we just put a little bit of effort on ensuring and validating that data going into the system is accurate as accurate as possible, then the resulting ending reporting on the other side is going to be much better. But don't wait for data to be perfect before you start using something. Navin, what do you see across, you know, both from a philosophical perspective, let's say without you off, but also what you see across your clients. Alright, I'll put on my legal ops philosopher hat, Phil. But, you know, I think that accuracy bit is always really interesting as it relates to AI, because that tends to be the biggest world. I'm afraid to use it because it's not 100% accurate and nothing ever is, it's just about the right use case to a certain extent. So I really think it's all about kind of being the right thing. So what I mean by that is the two worst things you can call a procurement team or really any is one a blocker, which I think Hal you said just now, but the other thing is cost centre. Nobody wants to be a cost centre and think about it in the context of I can be more of a strategic value driver, right? I've got all this information, I've built it through AI, through just looking at contracts and filling out stuff and a lot of the inaccuracy tends to be with legacy documentation because it has a lot of issues like your bad scans, handwriting and stuff like that. The more contemporary, natively digital documents actually lend themselves really, really well to AI analysis as well. But being that value driver, what does that really mean is that you can be a lot more proactive with your supplier contracts because you have that data. You can reassess, you can pivot, you can think about tariffs, for example, and how a lot of companies have had to restructure their entire supplier engagement relationships and you can't just do that because you feel like it, you have to know what contracts allow you to do them to begin with and what the penalties are and then evaluate if those penalties are worth paying because if you don't do that you wind up losing a lot more in the long run as well. And then, you know, it allows you to kind of share stories that are far more useful to both leadership and your other stakeholders across the organisation, like with negotiation. We didn't negotiate particularly hard and that's the end of the story, but rather where we had to pay penalties because we didn't complete signing a contract as quickly as possible. I think Hal, you gave that example about taking too long to sign and then the pricing changes, for example, or what specific issues there were. It lets you collect SLAs or measure SLAs a lot better, you know, because again, we negotiate these things so heavily in the contract and then we put it away and we forget about it. We put these vendors and then when you're evaluating vendors as well as suppliers, you don't have all the information to make the best assessment, which you can if you set up that infrastructure and you have the data into today's world in a way that we could not do in the past. Now, I know we've gone past that forty five minutes. I mean, I promise I'll get everybody out of here by the top of the hour, but it's a really rich conversation. I think we lost Kelly due to electrical problems and snowstorms in So I think that that was what was happening there. I do have a few questions before we do come in to the close. One of them, Linda, I'll ask you this first and Hal, feel free to jump in on top of it as well. We talk around increased data access. We talk around as much self-service as we can as possible, but what kind of guardrails or governance do we need to be thinking about to make sure that, you know, we're not letting everybody kind of run amok and do whatever they want? I guess I'll go first, Hal. Yeah. Okay. Source of truth, centralization, to ensure that data quality is coming in at the right point. Unlike how we at Boxwell, I don't share all of the sort of the supplier portfolio, like what I call book of business with our business partners, to have them like kind of self serve. I actually help them as a partner. We do meet with our business partners at least twice a year, at the beginning of the year, middle of the year, to align on strategies and strategic approaches to the supplier portfolio, I'll put it that way, but to make sure that they're also we're also aligning here's the additional value that we're also aligning and putting our efforts and resources around what's going to help us get to the company goals. Because we're all very, very busy, but are we busy for the right things and at the right things? So with that said, Hal, I'll turn it over to you. I think this information becomes invaluable around budget seasoning and planning budgets and helping business kind of figure out going forward, what kind of expenses they need to be prepared for, being able to have insights in terms of any kind of price increases and things like that. So having that data upfront and having those conversations with stakeholders, with the budget owners is key. You have to be careful for hallucination, right? If AI is hallucinating, so the human element has to be there to verify some of the data, to test some of the data, check some of the responses that are coming back. You just can't blindly trust what's coming back from the AI. They're getting better and the hallucinations are becoming less, but it's still something you have to consider. And then, having regular cadences with the legal department in terms of what we're seeing from clause acceptance, changes in contracts. And, is the red pen becoming less or becoming more or how are we addressing that? I think is important. And then the other key factor is sometimes having a business leader kind of come in and be like the final arbiter, right? There could be a difference of opinion between legal procurement, legal in the business. So, having someone at the C level kind of coming and being that arbiter to say, you know what, I think we're okay with that risk. Think we can move on. Navin, I want you to, you've worn many hats today in the last fifty four minutes. I'm going to ask you to wear another one and that's your legal hat. How can we as procurements make you feel comfortable as legal as we're getting more and more involved in things like, getting the data and extracting all the data from the contracts, making more decisions based on contracts, being more involved in red lining, and perhaps taking some of the things that you would ordinarily do as a legal team and bringing that into procurement? How can we make legal feel comfortable with that? Now I have to put on my dark legal hat because I'm really flexible. I want procurement to take on, I think, much as possible. I think definitely there are several elements. Some of it is roles based access. I would want to make sure that the right people have access to both the repository but also the review process as well versus just something that people can randomly pick up. And there's ways to handle that through the workflow as well. I want very clear escalation thresholds as well as far as when and this is something that varies from team to team but from a legal stance, I'd want to know that procurement is proactively escalating the right things to me and not just always relying on the way someone fills out an intake form to drive that escalation, right, because we've all seen it, I know Linda Chuan, you have too as well, where someone fills out a form and procurement looks at it, but really it should have also been escalated to legal. Knowing that that muscle is built or that we have some automation AI in place to kind of fulfil that escalation as well. I think the other piece of it really, and this is the most important across the board, but having effective audit trails. Because eventually something's going to go wrong and then I'll be looking for someone to blame. So, no, I can't, of course, but we'd want to know where the process broke down so that we can either train, yell at someone, or more likely to just fix the process and the workflow or bring in new automation. So I think those are really core pieces and the final piece of it really would be I would love to empower procurement teams to actually directly negotiate contracts because you think about especially things like statements of work and non disclosure agreements that we have, there's really no need for legal to look at this. Some attorneys will disagree with me on this, but I think this is again where AI can be really helpful to just not redline the document for you, but to just flag potential risk issues so that a procurement expert or procurement person can just exercise their judgment and go, yep, this should go to legal, or I can push back on it initially so that by the time the draft goes to Navin in legal, I'm getting the best first draft that I can redo instead of going back and forth between say me and Hal 15 times, before it goes back to, our supplier. So, yeah. Thanks. I've got a couple of quick fire questions in the next couple of minutes. We had a question that came in through Q and A. Linda, I'll ask this to you and then Hal, if you want to jump on as well. The question is, are you using AI to help generate commercial or legal arguments or even rights contracts? Are you using it more as an internal support tool? Oh, we're definitely using AI to help generate. I don't know about the debates and stuff, but definitely recommendations of what a new verbiage would be. You, Hal, I'll come to you on that one in a sec, but I have one quick follow-up. How do your suppliers react when, you know, if there's AI, if contractor are written or assisted by AI? I don't think they know at Yeah, this it's just like, hey, we're just providing the supplier here's our red lines. They don't know that it's really being recommended by AI tool in the back. Thanks Linda, Hal. So we haven't moved into the AI drafting contracts. It's still more of a review type of functionality. We're still using third party paper terms of the contracts that we're doing. So we haven't matured to a point where we're introducing our paper to the end suppliers. Okay, thanks. And then the last question, putting you on the spot a little bit, one common misconception that you had about AI enabled CLM and workflow associated with it before you implemented. Hal Wechter, I'll ask that to you first. So I had a concern that the AI would over edit or over complicate the review or the legal situation or come back with a million red lines. Actually it didn't, it was more moderated in terms of suggestions and things that it came back with. I pleasantly surprised that it wasn't as heavy a pen as I thought it would be initially. Great. And Linda Chuan? I had a misconception that AIs could get superheroes that can do whatever. And going forward, now I've come to realize that it's not going to be AI by itself. I think I've touched on this a little bit, it's AI and a humanoid. But we as people managers or management managers of a team that we will be managing more AI agents as sort of the tasks that they do. So yeah, so it's not just like, oh, you just kind of turn on AI and kind of just kind of goes. Right. It's a self-service. So just go. But he needs a judgment part. Great. Thank you. Well, again, last chance if you see that download button in chat, if you would like to get that report, now's the time to do it before we close-up here in thirty seconds or so. Unlocking contract data with CLM and AI from locked files to living intelligence. I wanna thank Hal Wechter, thanks Navin Mahavijiyan, thanks Linda Chuan for