Where the human stays in the loop when agents run procurement

Edited for readability from the episode audio. Filler and repetition removed, meaning unchanged.

00:11
PHILIP IDESON: Today, I'm delighted to be joined by Jeff Ostrander. Jeff is head of North American Basin Supply Chain for SLB. Our team at Art of Procurement has been delighted to work with Jeff on a number of occasions, and I'm looking forward to hearing his perspective on procurement's agentic future. First of all, Jeff, welcome to the show.

00:28
JEFFREY OSTRANDER: Hey, appreciate it. Thank you, Phil. Great to be here and excited for a Friday, at least in my side, when this thing is getting recorded.

00:36
PHILIP IDESON: Just for a little bit of context, can you give us a quick picture of your role? What you're responsible for and how procurement fits into the broader organization at SLB?

00:46
JEFFREY OSTRANDER: Absolutely. As you mentioned, I'm responsible for the North America Land Basin. It's part of what we call the geographic structure of how we look at supply chain and procurement. My role and my responsibilities in that entail is we run close to a $5 billion business, and I'm ultimately responsible for taking a lot of what we do at our global levels, matching that with what our ambitions are from our divisions, and figuring how we execute that on the ground. That's from the supplier identification, the supplier management, the logistics, the compliance organization, from the trade and compliance, our teams around import, export, as well as our regulatory, based on the types of businesses that we're in. If you look at it from a supply chain standard, we take the whole gambit. And like all things in supply chain, we have our hands on a lot of different levers because the business is so dynamic in the North America market. We have to play that, I wouldn't call it quarterback, but definitely someone that's in the huddle, trying to make sure we're doing the right decisions at the right time, which is always the fun part of the job.

01:52
PHILIP IDESON: And your team itself, how is it structured?

01:55
JEFFREY OSTRANDER: So we have taken, this was the decision we made back in 2022. We're structured in a way, if I look at my direct org structure and how it relates to the rest, we've adopted the model of having our sourcing organization separate from our supplier performance team or supplier management team. That's more on the, call it the traditional source to contract process. From there, we have another key decision that we made is our logistical outsourcing. We actually use a managed transportation model and partner with a third party that's really integrated into how we do our business. We have that structure. And then we have what we call the compliance organization, which is really centered around traded and adequate tariffs as well, trading and tariff organization. The import export was dealing with that, obviously, front and center of movement of goods internationally and outbound. And then, of course, our team that's looking after a lot of the regulatory components. But in traditional procurement, that key point of having someone that's identifying the suppliers from the sourcing organization, because the nature of the business that we work in, and again, like many companies, we have global category structures as well, but we also have what I would call a regional category point of view. We lean on our sourcing organization to wear that double hat. They're responsible for the supplier selection, but they're also responsible for developing the category, what we call category light model in terms of making sure that fits in the North America strategy and aligns with the global.

03:20
PHILIP IDESON: They organized around certain categories of expertise.

03:23
JEFFREY OSTRANDER: Absolutely. We've learned that lesson. It's not anything unique to us, but there's absolutely necessary to have people that are experts in their domain. They're comfortable with it. They understand the supply base. They understand the competitors. They understand the supplier base and the supplier competitors so that the sourcing organization understands that component. Now we've coupled that with our supplier performance and procurement to match and mimic. We have the exact same category structure on the supplier performance team. They're focused on our largest suppliers. Typical 80-20 rule applies, of course, way more dynamic in terms of how we select the suppliers we manage. But that link between having someone that's, let's say, chemical, chemical spend, sourcing organization works side by side with our supplier performance organization, and they feed each other in that loop as we develop our category light strategy, as well as just the day-to-day execution of supplier primaries, backups. And of course, when things all go haywire, where do we go and where do we go for what?

04:18
PHILIP IDESON: But then you have smaller teams that can face off against the same stakeholders within the business. The business are working with the same folks, whether it's from the sourcing, category management, supplier performance, as a small group.

04:29
JEFFREY OSTRANDER: And then the knock-on effect of that model is we actually co-develop category expertise, which then creates a talent pipeline in both domains. I can be in sourcing, or I can move to supplier performance, and back and forth. It allows us to not only move them left and right within the same category structure and same commodities, but then also cross commodity as they develop larger expertise. Then you get to a point where you're developing the future supply chain manager or sourcing manager that has now spent time in every category or commodity.

04:58
PHILIP IDESON: I know that culture is really important to you from a leadership perspective. Can you talk a little bit about the culture of the group that you're leading?

05:06
JEFFREY OSTRANDER: And I've had a unique time now. This is the group I started with when I joined the company 15 years ago. As I left and went inside the company into other domains and functions and other places around the world. But now I've come back to the same organization. I've got to see the culture I started with and what I came back to and where we are today. Culture plays a fundamental piece of how we run the teams, how we select the people that are part of that team, and how we develop that as from a culture standpoint. If I singular define what I say the culture is of the North America Land Supply Organization, I'd almost put us in the evangelist role because we are a unique place where this basin in particular, the North America operating market, U.S., Canada, the operating business model is slightly different than the way we operate in all the rest of the world. The suppliers that we're working with, the dynamics of the market, the hunger and pace and scale is very different from my counterparts. The culture has to fit that. The team has to be champions of what they do. They have to be inquisitive of what the business needs, and they have to be really energized to go after it. We look for those profiles and we build our culture on finding the right folks that can fit that model that are hungry to go out and research and find out ideas and problem solve and willing to engage with the business on what they're actually trying to achieve. Because again, we come from a company that's from engineers whose mindset is really around execution on behalf of the clients. And sometimes the supply chain piece gets overlooked. We play our role of, okay, that's where you want to get to. There's this way, road A, road B or road C. Each one comes with a different risk and different speed. Each one comes with a different value lever, and we can do any one of those. But we have to be that referee in some cases to say this is the right call for what we're trying to do.

06:56
PHILIP IDESON: But taking the true business partnership approach, not necessarily saying yes to everything, because you still own that responsibility to the business ultimately. But you're going in there with not throwing process and procedure, but more about what's the business outcome that you're wanting to achieve.

07:12
JEFFREY OSTRANDER: And it's been a journey we're no different. We're a company that's built with a lot of processes and procedures. I had the luxury of spending time in our global business services organization that's founded on having strong process, procedures, transactional aspects. And there's an ingrained thought on this that cascades through policies and controls and this stuff. And now I think we're in a place where we spend a lot of our time making sure that actually adds value. The whole premise, when I look at the North America market in particular, the guiding force, and it's hiding behind me We call it the now experience. But it's, we need to rethink what it is we're trying to accomplish here and what's actually adding value. And over time, the longer the organization's been around, our organization, Schlumberger, now SLB, just celebrated a hundred years. We've been around a very long time. We've been in this space a long time. Supply chain obviously not been alive in the company that many years, but we have a long history. And with that long history, a lot of baggage and a lot of baggage of policies and procedures that fit the time, but need to evolve with the way the org's changed and just the way the reality and the ecosystem has changed So we spend a lot of time trying to make it easier to work with, to show that we can drive the value in a much better way for the company and not have any exorbitant risk associated with that. But the right risk is a good way to think about it.

08:33
PHILIP IDESON: Now, this podcast series is all about the agentic procurement future. I'd love if you could just share a little bit about where are you on your journey with agentic AI? Are you at the beginning stage, just really exploring what's possible? Are you you've been putting it into practice and you're have been going that necessarily full steam ahead, but certainly beyond the, I wonder what this can do for us stage.

08:59
JEFFREY OSTRANDER: I think, and this is a, this is core to who the company is and we're not opposed to risk by any means, but we're also pragmatic. I think the easiest way to articulate it, we've left the starting block for sure. We're in the decision-making of how far the race we're going to want to run. Like what's the distance we're willing to go here? We're off to the races for sure. We got a lot of hands in different areas and playing with different things. We started very early in our journeys, of course, long-term partnerships with Microsoft. We've tapped into other models as well. You have a long history of building automations, building machine learning, and now agentic comes around and opens up doors that we didn't have before. We've put our hands in that. We started in sourcing for sure, working with the end-to-end engagement from a sourcing standpoint, the creation of content and the ability to pull together a lot of resources, both baseline data, baseline spend, aggregating gaps in our data to make opportunities for tendering more realistic, looking at planning, of course, cycles and those kinds of things, which is the most complicated one to actually get your hands around because of data. Legal and contracts, scouring our legal language, making sure it fits when issues come up or challenges come up in the marketplace that have a certain context, whether it's the tariffs recently or changes in commodity prices and fuel costs, how many contracts is that embedded in? What is the risk to the business? What are the financial risks to the business? These are all things that you can leverage that technology. Tail procurement. When you're asking people to make logical decisions, probably the most easiest one from a closed-loop standpoint is getting closed-loop, end-to-end tail procurement where you have some defined guidelines of only work with these suppliers. Great solution for these toolkits. And then more recently, and one that I think is really exciting, at least from what I've seen on pain points, is really around that back-end administrative cycle. I think more from compliance, freight pay, assurance, invoice comes in. Does it match? Does it match up against the tariff policies? Is the billing correct? And if it isn't correct, can the agent work directly with the submitter and get it fixed so you're not sending it to a ticket or a person and they have to go do the work? You can go do that simultaneously across a thousand errors where before you churn through it So these things are fast-moving, quick, practical, and we've got our hands on all of them. From a supply chain standpoint. From a company standpoint, we're way past that world We launched the very first one in the oil and gas industry. We partnered from the fact that we just had so much data, so much pieces of the story already pre-built, and we had gone early to the cloud with our system on the Lumi platform and where we house a lot of our data and then put on top of it what we call Tela, the assistant. A lot of our clients now can actually interact with their data in native language, which is pretty phenomenal in our industry. And supply chain is no different. The amount of time and the data and stuff required to make decisions was so complicated. AI is removing those barriers and the speed to get to that data. It's changing, I think, people's perception of what the world will look like going forward.

12:27
PHILIP IDESON: It sounds like SLB is pretty forward-thinking in terms of AI adoption. As you talked about, even before AI, when you looked at automation and coming into it so that AI itself wasn't a brand new thing where there was no history of thought that went into its application. How did you think about like, okay, where do I start? You walked through a lot of different use cases at work right now. Was there a moment where you thought, wow, okay, I have seen this. Now I can see the future of what AI may do for me. And these are therefore all these different use cases, which I think are possible that may not have been possible before.

13:10
JEFFREY OSTRANDER: I think as a company, we were smart enough to start to really ring fence some folks to take hard looks at this. The IT and infrastructure and really the client side of the data and what that means to the industry. They really went down that path very early and success to them on all that they've accomplished. I think along with them supply and what we call global supply, which is really a combination of traditional procurement and sourcing and supply chain management, coupled with maintenance and what I would call your manufacturing element So bring all those forces together. We call that global supply at the company. We were smart enough to ring fence some really good experts that had their ring on the pulse of these early signals of AI, machine learning and infrastructure of what we deal with from a supply chain system standpoint. We're like any other company, a lot of different tools we use, self-built, external, tied to our ERP, lots of those kinds of platforms. But we put really smart people that really understood what was happening outside of SLB. What's the flow? Sent them to spend a lot of time at conferences, a lot of time engaged with working with suppliers. What are they up to? What are they working on? How does that fit? And building internal cases that we think there's some opportunity here and then the willingness to then invest. And like all times from a company standpoint, we've been very good about targeting, I wouldn't call the big box suppliers, but the ones in the middle or the young, innovative ones that have something that's really pretty cool and are dynamic and are willing to work with us to build what that might be like the first mover, those elements. And the challenge now is there's so many. You have to have people constantly in contact out there and having conversations and learning and seeing the demos and how does it translate and asking questions and educating themselves on what's happening in that space. And your event in LA was a great example of that. I learned a lot coming out of Calus because honestly, there was a lot of the same reality of this is moving so fast. They're not even picking a player yet They're bouncing between multiple. And I think that is the state of affairs for everyone in the industry. It's moving so quick. Our business cases are changing daily. Once we solve something, we open up another potential opportunity that cascades. You need to have people constantly in that space. It's not something you pick up on the weekend for an hour and get your head around it. You better start living it. You better start being interested in it. And you better start finding the people in your company that want to be in that space because they will be self-motivated to go and chase these things.

15:40
PHILIP IDESON: How have you, because it leads to an interesting question. You said it's not something you're just going to spend an hour trying to understand on the weekend. I think that if you try and do that, you get overwhelmed. How have you managed the overwhelm of all the different potential opportunities? Because I think that's where we are right now is that there are so many different potential ways that you can go that it can become overwhelming. When it becomes overwhelming, it can be debilitating. I'd love to know how your mental model of trying to deal with all these different possibilities.

16:12
JEFFREY OSTRANDER: So my approach, I guess, is mine. The way I've always looked at it is the perspective of number one is I like multiple points of use and I seek it out. I spend a lot of time, there's some really great sites and stuff that I use and follow and folks that I follow these days just to see what they're talking about, what they're up to and trust a bit over time, their opinions on certain topics. And of course, that drives some of my decision making. I use AI a lot to help scout out things or post challenges and where they're seeing solutions being marketed along those. That nails down some of my searching. And then a lot maybe it's a culture of the company that we're in, but we do a lot of discussions We're a company of a lot of dialogue and challenging each other and posing opportunities and the network within the company itself is quite strong. People that have their hands on these different areas, whether it's from our IT organization or even from supply, that you can have these, this is what I found, this is what I found. What did you see it do? Okay, was it capable of that? Is it not capable of that? Where did you see some risk? And you can have some real theoretic conversations and get to a point where you say what, I think between one or two or three of these companies, we probably have something that's in the right box. Let's engage a bit and have a conversation and see what that might look like if we go one step further and actually do a pilot or pick a location or pick a process and throw it over the fence and see what comes back. Knowing that we're going to stumble or we're going to go a bit slower sometimes or we're going to get it wrong and we're going to course correct. And I think that's the mindset people should have today. You're going to have to get your foot in the water. You're going to have to test it out. You're going to have to decide what you're willing to take some risk on and know that you might not get it right the first time, but you'll probably have a much better success the second time around that you'll learn from it. And I think that's the mindset you're going to have to be in. It's just too dynamic to not have your hand at least partially on that.

18:08
PHILIP IDESON: There's so many learnings you can take from doing something that even when you're just playing, say, playing around, that's not the right way of saying it. But when you're using the Frontier models and the Frontier models are changing so quickly that even the way of doing something and achieving something using a Frontier model today may be different than how you go and use it in two months time or three months time or a week's time because there's an update to a Frontier model. But you can never take away the learning that you take from doing it the first time that you can then apply to the others. Don't look at it as wasted time or I'm just going to wait until the model evolves. I think that's really important.

18:44
JEFFREY OSTRANDER: You're spot on. I think that, and this has happened, we've all been through enough iterations in our life of seeing these transformative cycles. And this is obviously a new one. It will become part of what we do, of course. I think it's got sticking power, for sure. There's just too much upside opportunity across all of our things that we do. And like I said, the amount of folks that maybe have been operating not as the high potential crew, but maybe they're right below that, or they haven't figured out how to work better in certain areas of their career, this generates a massive value for them to upskill themselves and get to a point where they're actually contributing far beyond. And to your point, I think the narrative of what's happening for each one of these streams and companies and where they're attacking, and some are very niche If you look and you spend any time, you'll find folks that only focus on this problem. Now the future state, and we all know how it works some of the small ones may get digested by somebody else as they build out their value proposition. But those little niche guys, phenomenal to invest your time on. And phenomenal to have the conversation and say, okay, how'd you get there? What was your logic? What are you trying to, where do you see this form ending? Many of them are open to have those types of conversation at this point They're trying to build out their own value proposition as well, and they need people to help them build it. It's an interesting and worth effort to put yourself out there and dedicate time. I know it's hard to, we talked work-life balance. I was joking internally with some of the groups that I have to mentor internally in SLB and I call it the work-work balance. You've got to carve parts of your work balance out as well for work-related stuff like this and discovery for sure.

20:25
PHILIP IDESON: As you've gone through the journey of adoption and pushing adoption through the team, has there been a particular decision or decisions that have been met with a lot of skepticism where you've had to make a case where others may not have been quite so on board or they were fearful of what you were pushing? I'd love to think about how you, because I think this is going to be something that as you talk about experimentation and risk, there's some elements of taking personal risk, but there's also some elements of you have to experiment to develop as we talked about before. And that's going to mean things are going to be met with skepticism and sometimes it can be easier not to do it because you're afraid of the pushback. Have you found that you've had that on your journey so far?

21:09
JEFFREY OSTRANDER: I think in general, we have to go into it knowing that this will be a change management exercise across the board. Everyone on the team will have a different part of that journey. Everyone will advance at a different pace and others are going to adopt it because they love the idea of it and they're going to invest mentally all in. And we see this not just in the domain of supply chain, but a lot of other things that people harness their energy around. If it's something that really motivates them, they will go full speed. Other parts of the team will not. And that's a fact and that's a reality. And it's okay at the end of the day. I think most people are scared because what they hear is the larger narrative. Okay, AI is going to replace all this jobs. People aren't needed anymore. And I think that's just coming from not having that next layer of understanding of what it's actually doing and really understanding where the human fits in that loop We always call it human in the loop is the terminology we use the most. I think I see a world where that's still going to be the case. I think the human loop will be there for a while because there's just some, at least I haven't been sold yet on how it's going to work end to end seamlessly by itself in a full end to end workflow. Parts of the workflow, I think for sure they're going to have to just take over because they're faster, they're quicker, they can connect dots in an easier way. And if we ring fence it or put them in the right guardrail, then that's going to be a phenomenal improvement for everybody's lives. But I think most people come from a fear of what it means versus just a willingness to understand it. And then once you understand it, where does it fit? And then once you understand where it fits, where else can you push the boundaries to expand what it can do?

22:47
PHILIP IDESON: And I want to come back to the human in the loop for procurement processes in a second. But knowing your background as well, you talked about the fact you ran business services in Columbia. I look at this as being, from a change management perspective, it's not really any different than moving roles, tasks, activities to shared service centers than it is now you are moving them to a faceless agent as opposed to somebody who lives and works in another country. But the underlying fear of what's this going to do for my future and my job is the same fear. It's just applied to an AI agent rather than somebody that's working in a global business center in a different country.

23:34
JEFFREY OSTRANDER: And it's a good example, actually, because, again, we look at when I remember the first transition when we moved things into these hubs in various parts around the world, and that was a big issue. Okay, I don't have my guy down the hallway anymore. How is this going to work? It's an arm length now. It's a phone call. I think COVID threw that on hyperdrive. We all got used to working remote and that barrier disappeared really quick. But the fundamental piece was still there. And then it was more around the roles that they're in Okay, I'm in a procurement role outside of the main operations. I'm in a hub. I'm getting requests in. I'm processing those requests. I'm interacting with the supply base. It's predefined. I'm asking them for quotations for parts that aren't in our catalogs or fixed prices. I'm coming back with an answer. I'm cutting the PO and it's off the door. If you understand where AI is, there's a great opportunity there to actually route a lot of this through that mechanism and out into the supply base that is capable of receiving it. That interaction, and I'm sure over time, AI will get better. The voice systems will get stronger and the interactive will get stronger. And you'll get to a point where they can interact with both the tech-savvy supplier as well as the not tech-savvy suppliers to have that conversation and get quotes and deliver an executed base. Now, if you're a procurement specialist, you see that as, oh my God, that's not great. This is my career point. I'm worried. Whereas I think the real reality would, well, if you're not doing that part of the process, now you can actually understand what's happening from a procurement consumption standpoint. Am I consuming the right amount of goods? Is there an opportunity to challenge the specifications? Is there a challenge to look at actually how the demand signals are being generated or going deeper into the supply chain beyond the transactional workflow receiving an order, processing an order, typing an order, releasing an order to really understand what the client and customers are doing and see what that means to a bigger picture. That value proposition and cross-training folks that are doing those more administrative tasks opens a huge amount of workforce to the team to really get to the areas that unfortunately, because of those tasks, they haven't been allowed to dig deep into. I think there's the huge upside. I think when we talk about efficiency and just capabilities of companies that have that an operating model, it's going to unlock massive amount of horsepower that we have not seen for a long time.

25:47
PHILIP IDESON: And with the shift to business service models, this hasn't happened before. We're playing the same. It's the same playbook because that's the same value proposition. It's not a hypothesis that you don't fear your job because there's going to be all this additional work for you to do that's going to help you advance your role, your impact, what your organization can do, because that's what's happened in the past. I think there should be some level of confidence that there's already a defined path to walk in terms of how to take advantage of the change in front of you. You talked about not necessarily trusting an agent, an AI agent to act without a human in the loop. What do you feel needs to be in place for you to be confident for an agent to act without a human in the loop?

26:43
JEFFREY OSTRANDER: And again, I'll limit it to my exposure so far on the stuff that we've been looking at and testing and these kinds of things. I think a couple of things from a human in the loop perspective, especially when we get into actually, this is where you get into, okay, where does this actually go and whose role is it? When I think of compliance, when I think of things like even purchasing execution decisions, interacting commercially with certain suppliers, how much autonomy do you give to the agent? And what can it do or can't do? And human in the loop for me becomes critical in some of these decision-makings because again, depending, and as we talked before, depending how you articulate to the agent of the boundaries of the agent, what you're willing or allowing it to go after This can get you exposed to discussions around if you're opening it up to select new suppliers, now you have a risk. You're working with existing suppliers. Are you making purchasing decisions based on what criteria? And what feedback loop does that AI agent have in terms of just visibility to supply chain Supplier can commit to a price, but they might not have capacity or the capacity doesn't meet the lead time or the lead time you have doesn't match reality. You can give a lot of frame for the AI to work with it, but without a human in the loop, you can make a lot of incorrect calls based on what you're asking it to do and the data issue will come back and haunt. Now, again, there's always this chicken egg. I think if you were to have the AI discussion nine months ago, people would say, well, I need to clean my data before I can use AI. Otherwise, I'm not going to get good data and answers. That's partially true. I think when you get into much more complex scenarios like planning and execution and resupply and all that fun stuff, data then becomes obviously massively important. And as we know, most organizations are not clean. They don't have a perfect answer in the crystal ball on all those elements. They have a lot of gaps. I think AI will help bridge some of those gaps by looking into other data sets that exist, but maybe we haven't connected it internally today or maybe it's already happened. Visibility from your logistics providers on quantities or shipping POs and everything can be put back together will bridge some of those gaps. But I think the human in the loop perspective for me is going to come more around, okay, if the AI agent makes a decision commercially, who ultimately made that decision? No longer attached to a person. Who legally is responsible for that purchase? Who's legally responsible for the issues that may arise from that purchase? There's a lot of that fundamental that goes then beyond supply chain. You've got HR, you've got legal, you've got all these folks involved in terms of what that means. Is the person that coded responsible? Is the person that allowed that flow to be automated responsible? Who's responsible when the supplier says that's not what we agreed to? There's a lot of that stuff that I can think is inherently going to be challenging for people to solve. I think for the traditional, again, in our world where our supply chain is extremely dynamic and the types of goods and services that we provide are very dynamic and not as clean as let's say a consumer goods or something that's a bit more highly data integrated.

29:47
PHILIP IDESON: A bit more material based.

29:48
JEFFREY OSTRANDER: Be the first movers, I think on that and connect the dots end to end quicker than I think the ones that have more of a service element or just a supply base that is highly dispersed with different levels of technology advancement. That's where there'll be some breakdown. And I use the example all the time in oil field services as an example. You've got machine parts suppliers. Most of these guys that do phenomenal work are not in front of the computer. And so that is a gap. AI is not going to be able to solve that and connect that part of the chain.

30:20
PHILIP IDESON: You have to have a supply chain that is willing and able to interact with the way that you want to interact with them.

30:26
JEFFREY OSTRANDER: Again, I think there's a lot of, and not to belittle, I think there's tremendous opportunity. There's a ton of opportunity, even in those supply chains to augment certain paths. Me as a comfort level, because I've just seen so much iterations of it, I'm not 100% sold of letting the reins off yet. That will probably change over time. The more confident you get in the model's decision making based on experts feeding it, I think that'll get closer to a reality where certain parts you say what, this is landing at a high enough percentage. It's probably better than the human element, probably making similar mistakes. It's better off there than it would be leaving it.

31:03
PHILIP IDESON: It's interesting how sometimes it feels like we have a higher, or should I say a lower tolerance in terms of mistakes and the impact of mistakes when a human makes them than we on a, when an agent does, probably because we feel like we have more control in fixing the challenge because we understand how it happened. Whereas we don't sometimes do that with agents yet. But as you said, I think that with experience and with continued development of everything that goes around the agents, then some of those concerns will probably be less of a concern over time. Now, I know it's about time to wrap up, but I've got two last questions that I'd love to just get your perspective on. First of all, it's just the journey that you've been on with implementing agentic AI across procurement supply chain so far. Is there any advice that you'd like to leave listeners with from your experiences?

32:07
JEFFREY OSTRANDER: I think, like I said before, I think there's a couple of things. For those organizations that are on the early road and trying to figure out where they step into this stuff, it's worth the time to invest creating some level of education across the board Like I said earlier, a lot of organizations are going to have those handful of folks that are very motivated in this space, and they're going to go full at it, whether it's at work or out of work, they're just, they're interested. And those are great. And you need to harness those. You need to cut them loose on going out and figuring out all the great players to work with. But beneath that, the change management story, you've got to invest the time and build that grassroots understanding of what's going on around them. We took it internally. We took a lot of curation of content, worked with a lot of partners to build it and give some general education to the full population. And we did it intentionally knowing that this was on its way, bring everybody up together. At least there's a fundamental understanding. And then from there, grew it even further to help develop those SMEs and develop those expertise that other people that are motivated by it, but then also bring the leadership team up on, okay, these are the, all the capabilities that exist or could exist in base case and realistic real world scenarios of where we're already doing it. People's minds can think that way, because I think that's the biggest piece as managers and as leaders, we have to start to connect those strategy dots of where we are and what, where we could be in making that happen. I think that's the most important part.

33:32
PHILIP IDESON: Thank you so much. And the last question, which is really to get out your crystal ball, which is probably the hardest thing to do right now. I promise you, we won't come back to you in three years and hold you to this, but if you had to place a flag in the ground, what are some of the things that you think would be true about procurement in three years and not necessarily obvious today?

33:54
JEFFREY OSTRANDER: Three years seems like a long time, but it's not that long. I think AI is going to, I think human will remain in the loop for all those folks that are worried about AI is going to change and disrupt in three years. I don't think we're going to get that quick and fast in procurement, at least in our industry. I think there's just too much dynamics that it's going to take longer to solve and too many partners that are not going to be enabled, but AI will change fundamentally how we, how we work. It already has. If anyone on this phone call or on this podcast hasn't started this world already, your life is now, you have 10,000 more horsepower than you did yesterday because you can aggregate data across so many sources that you didn't have access internal or external, by the way, that you just, it would take you quite a bit of an effort and you can now do this in an afternoon or in some cases minutes, rather than what you've been doing over your time of investing a tremendous amount of horsepower. That's one thing. The other part that I think is really important is that some of these processes and services, supply chains and the human decision-making in there, this is going to grow at scale. You're going to see some opportunities and some great, I'm sure, public case studies where people are going to release fundamental changes to how they look at the business. And I know there's some very early adopters. There's some very people that had unsuccessful attempts, which is not a bad thing, by the way, to know what didn't work. What I'd be careful and fearful of folks is to not use those as, okay, we shouldn't do it. We don't want to do what they did because they had a mistake. You're going to have to iterate because I think the companies that figure this out, the companies that figure out how to put it into the workflow, rather than looking at it as like a replacement, I think are going to have tremendous upside value. And I think this is where our mind has been is we believe at some point, some of these workflows can be automated from end to end and agentic AI will be better served for it. What's coming with that expertise, where I fundamentally really enjoy the interactions is when you start to bring in the native language engagement, folks that are unfamiliar with an ERP or unfamiliar with the reporting, you can use that interactive capabilities and agentic and do that to let you in native language, ask questions from your perspective, what you're trying to answer. And it's curating that information back is just an absolute game changer for folks that are not as familiar. And I think that's from a supply chain professional. Our biggest challenge has always been educating people on supply chain. It's always the black box. This will help, I think, bridge that gap. A lot of folks that are unfamiliar with supply chain and the complexities and probably engage and allow us to engage at a level that we've never done, which is really great.

36:29
PHILIP IDESON: The UX doesn't matter anymore. Or the UX meets them where they want to be met and where they work and how they're comfortable and in the language that they're comfortable in working as well when you're working across multinational organizations too.

36:42
JEFFREY OSTRANDER: Exactly. And I think you may find at some point in the future, Philip, is you're going to have people that are never interacting in the ERP because they don't have to.

36:51
PHILIP IDESON: Well, Jeff, I want to thank you so much for joining me on the podcast today. Just sharing some of your experiences, some of being open in what you're working on as well. It's always a pleasure to chat and I'm looking forward to check back in as you continue on the journey at SLB.

37:06
JEFFREY OSTRANDER: Appreciate it. Thanks. Love the discussion. Looking forward to the next one.

37:10
PHILIP IDESON: All right. Thanks a lot, Jeff. Take care.