Ep 03: The Agentic Procurement Diaries: Boring Enough and Safe Enough: When AI Gets the Keys

Hi there, my name is Philip Eidsen and welcome to the Agentic Procurement Diaries to show where we explore what procurement could become in an age of Agentic AI. The Agentic Procurement Diaries is presented in partnership with Zikus, a leader in Agentic AI for autonomous procurement. If you are a CPO navigating this shift, this show is for you. The truth is that nobody has all the answers yet, none of the most forward thinking leaders, wrong learning, experimenting and figuring out what this technology could mean for our teams and our function. In each episode, you'll hear from CPOs about what they're experimenting with, what they're learning, the challenges they're encountering and how they're preparing their organisations for what comes next. These are conversations about the journey told by the people taking it. And today's show I'm delighted to welcome Riza Symbol, Riza is the director of supply chain management for TransAlta, a publicly traded power generator based in Canada, as customers across Canada, in the United States and in Western Australia. Well, first of all, Riza, thank you so much for joining me on the show today. Hi, Phil, and hi everyone, and I'm happy to be here. Well, it's our pleasure and let's kick things off just by understanding a little bit about your role. If you could give us a quick picture of what your role is and what you're responsible for and then how procurement fits into the broader organisation at TransAlta, that would be really helpful. For sure. So first of all, and I'll quickly touch on something that you already mentioned, but just to give them better context. So TransAlta, I leave the supply chain function at TransAlta. And TransAlta is one of Canada's biggest power generators. And we've been generating electricity for more than 115 years. So we have a very diversified portfolio of more than 90 generating facilities in Canada, in Australia, in the States. And the assets they spend hydro, solar, gas and battery energy storage. So that's a lot. And then supply chain in TransAlta is not a back office function. So we are an operational one. So we support the full spectrum of everything that the company needs. From the day-to-day procurement out to just the projects, IT corporate naming. So we purchase older equipment that will require for the growth of the company. We do complex negotiations with OEMs, we do everything. And the way I look at supply chain and what my team does, and I do it from three lenses. It is reliability because we're power generator, it is speed and it's cost. And those can really pull you in different directions. So my job is finding the right balance for the asset and for the risk. And how is your team structured? So when you think about your team specifically, what does that look like? So I have roughly 30 people across buying and strategic sourcing. And we support the entire company, as I said. So our supply chain is centralized, but the entire team is not sitting together. We have some people in the corporate office here, but also at sites across Canada and Australia. So the buyers, they're the closest to the business. They keep the company moving. They're doing the purchase orders, the outages, the expediting. And then sourcing specialists, they're more on the complex site. So they're doing the category strategies. They work with major suppliers. They're doing supplier performance management. But that said, I don't believe in high walls between strategic and operational procurement. Because strategic work, we have to understand the operational consequences. And then the operational work needs to understand commercial context. And then for the last probably three and a half years, we've been moving away from being a summer over a purchase order factory into more of a strategic function, so we're doubling down on category management, supply relationships, and early engagement in business decisions. How do you manage those handles? Because I think it's really important what you said is that you can get into these silos where somebody who might be in a category management lead, for example, things. Well, I'm responsible for category, so I'm responsible for the strategy. I'm not too concerned about downstream. But of course, the decisions that they make impact downstream. And if you're not careful, you can create this world where they're not connected. So how do you think about ensuring that everybody's connected, so that everybody kind of does understand? Is it one role that does multiple facets? Or do you have other ways of making sure that there's connections between folks who are low core in the facilities, and those that are in head office? This is an excellent question, because I've seen that before. What you described, solid approach, my turf, staying in my lane. So what do we do? We do the cradle to crave. And we also flex people across the portfolio. Because when you think about it, business doesn't really come in neat, operational boxes, or when the business looks at the orchard, it don't really carry as much. They just want things to be delivered on time, on budget, on these days, even ahead of time. So my role is to make sure I understand the full picture, explain the why to the team, and then give them the opportunity to deliver what's required for the business, regardless of their title. So it's not about the title, it's about that work needs to be done. Yeah, the outputs, like their experience of working with procurement as a whole. Now, this podcast series, we call it the agentic procurement diaries, to really focus on agentic AI. And I'd love if you could share a little bit about where you are on your journey with agentic AI. Phil, we're early. Yeah. We're really early, but we're seriously exploring. And the assistive AI through co-pilot and other applications, it's been around for us for the last several years. So we're using co-pilot and other applications for contract reviews, for the first drafts, like scenario planning, maybe even getting to the better question sooner. But agentic AI is a different ballgame altogether, and this is where the value comes. So we are now at the point when we're focusing on practical use cases. So we are, for this year, we're identified to areas. And the first one is requisition to purchase order automation. And the second one is contract and purchase order leakage. So the pilots in my view, they have to be small enough that we can control them. But also meaningful enough that we can measure the input on the work. So for us, once we identify those areas where we want to pilot, we will also want to be cognizant of the impact of the company. Because Translata, we've come through a major ERP transformation just last year. And that's been quite a journey for the last two and a half years. And also the business, we're in a very active growth phase. So we're growing through acquisitions and through development. So the internal capacity is at the premium. And personally, just to share something with you, I'm trying to avoid this whole AI formal. Because everyone on their dog is talking about AI. But AI is everywhere, but the skilled results are not. So I wanted to start with the business case, where I would define the business outcome that is clear, well structured, and then decide where the answer belongs on the platform, let's say SAP and Jewel, or if we want to consider a point solution. So this is where we're at right now. OK, was you are looking at the different ways that you could deploy all the different use cases for a gender AI? What were some of the considerations that went into selecting you know, REC to PO, the contract and PO week, you're just being kind of your two test cases? So initially, we were thinking, what are the cases where we have a workflow that is high volume that has defined rules and that no one really likes doing? And those are two cases where they fit the profile exactly, we have high volume, we have high volume of transactions, we have rules based, we know exactly what constitutes a good acquisition, what needs to go on their purchase order. Let's say we have defined terms and conditions, so the agent can compare the invoice for the leakage against the invoice. So we were using those criteria. High volume, people don't like the workflow. It is potentially reversible if the agent makes a mistake and it's rules based. So those for criteria. And was there a specific moment where you felt that a gender AI is something that can really help us in these areas? Because these have obviously been pain points that you've had for a long time. And I can certainly attest to even the REC to PO process. Lots of pains, lots of pains for us as procurement organization, lots of pains for the rest of the business. What moment did you think you know what I think that a gender AI can be the solution to that problem? It wasn't necessarily the tri-readable. to pilots. And it wasn't really a lightning bolt. It was just a random Tuesday. And we had a technical issue with one of our OEMs. And we needed to find a specific technical piece of information. And we knew that it existed somewhere. It could have been on an old drawing, in the contract, in the email chain. But the problem was finding that piece. And I was watching a lot of smart people, days and days out, looking across documents, opening folders. And this is what Tristan lies that to me. So the information that you cannot retrieve last enough might not as well exist. And in negotiation, when I reflect back on what my team does, sometimes the value of information is partly about speed. And if you find an answer a week later, you might have already lost leverage or time. So that's why I started thinking about, how can AI support us on that joining? So this was one angle. And the other one is, as we're getting bigger is the company, we cannot continue extending and expanding the support structure to fit the growth of the company. So we need to find a way to create capacity in my team because supply chain talent is scarce. So we can remove and strip out all this transactional and repetitive work. So people can actually focus on things that moved the needle that are truly valued at. So that's what that was the moment. Yeah, we're all under this pressure of doing more with less. And how do you kind of figure out where the efficiencies are to do that? Now, as you started to explore Agenda AI, did you feel that or did you find those any pushback or unexpected pushback to plans? Or was this something that you know, it's like, why have we been waiting so long? I'd love to know kind of the resistance or the encouragement that you may have come across as you start for these plans in process. The pushback was definitely there. And for the most part, it was reasonable. And the answer, sorry, the questions that were being asked, can we trust the answer? What data is it using? What about confidentiality, cyber security, the data access? Are we even automating a bad process? And then on the people's side, of course, does it eventually mean fewer jobs? But when you think about it, there is also a business case and a business case question when it comes to AI, because AI demo might sound really compelling. But every leader wants to know what would be the impact on their PNL, on the productivity, on the risk, and it is all fair. So the way I think about it, if I cannot connect the use case to a measurable business outcome, that it's probably not ready for investment. How do you think about the role of the human in partnership with the AI? Do you feel as though there's a world where, for certain tasks, they can entirely be done by an agent, an AI agent, and there's not really a human in the loop. There always needs to be a human because of that trust factor. So the task that I would typically use is, if this is low risk, high volume, rules based and reversible. And I want all four. So this could include routing of the request, checking completeness, exactly the requisition to purchase orders, the leakage and so on. So monitoring this tail span transactions that seemingly will never end. So this is where I think we can have the full AI automation. But then where I would not even think about at this point, handing anything over to AI, if those activities fall into this other box, is when we have an activity that can create a new material obligation, if this is a major negotiation, if this is a critical supplier, or if there is a decision with a safety implication. Because when you think about it at the strategic end, AI should make the human better, but not remove the human. But at the transactional, once we make those transactions boring enough and safe enough, I think the human can only remain as an oversight and not in the loop. Yeah. When you say a processor, an action, or a task being reversible, could you just expand a little bit on what you mean by reversible? Okay. The reversible decision is if I cannot unwind the decision made by the AI, then a human needs to be the one approving it. Because I could never go to an auditor or an executive and say the system did it. Yeah. So it needs to be easily reversible. Yeah. So for you, you know, you're as accountable for the actions taken by an AI agent, as much as you're accountable for the actions that might be taken by a member of your team. And so therefore, you know, that's kind of your level of, you have, you want oversight accordingly. You know, where do you feel comfortable with an agent doing something where, where do you not necessarily feel comfortable with an agent doing something based on that accountability? 100%. Yeah. Is, how do you feel that your team are embracing the use of a genetic AI? You know, I think you mentioned the fact that we should be using these technologies to empower our teams, to give them the ability to be able to do more with the time that they have, you know, to work on activities that are more strategic. And of course, there's always fear about what it means and some of the tasks because the role might look very different in the future. Is this something that you've found when you've, you know, socialized the, the kind of direction of travel? Or I'd love to kind of hear what you've, what you've found from the team themselves. And again, we're still very early. So the pilots, the team haven't seen the results of those pilots yet. And I'm thinking that there is definitely an element of fear because to your point, what will their role and their job look like in a year and two years and three years from now. But the way I'm thinking about it, the strategic end of supply chain, it will become more human because we will automate a lot of the transactional. So people will have time and capacity to focus on things that, first of all, they probably enjoy doing more than the transactional and repetitive can also learn more and will be more meaningful for their own career development. So this is on their team side. But then the way I'm thinking about the whole, the job profile of the future supply chain person, I think we need to be very cognizant of what are the changes that we're making because if we're going to automate much of the junior work, how do we build senior judgment? Because we cannot automate the so-called apprenticeship out of procurement. And then wonder, where is the next generation of commercial capability comes from? And that means that we really need to be more deliberate about building this kind of commercial acumen, the side, the supplier exposure, the decision skills, like the AI literacy, or maybe even hiring for a different mix of capabilities. So do you feel that that could also mean not necessarily selecting every single use case that could lean itself to being used by AI agents because you actually want to create a pipeline of talent who has that experience of doing the same thing. So not just because you could automate it off, because you could use an agent steer it, doesn't necessarily mean that you should do as you're thinking about kind of the talent pipeline and the exposure and experiences you want to give people. I really like how you summarize that, because in my view, the goal should not be to have the most of AI. The goal needs to be making better business decisions faster and with the right controls. But it cannot be one-size-fits-all. Right. So I want to just leave with, I think, a final question to ask you to look into your crystal ball a little bit. And I think this is probably the hardest question because we don't really know where things are going to be in 12 months time. Nevermind, you know, three is our five years. But is there something that you believe that would be true about procurement in, let's say, three years time that may not be obvious to us all today? This is an excellent question and something that I've been thinking about a lot, because even to your point, everything in procurement changes only weekly basis. So when I'm looking at my crystal ball of the three things that I'm thinking about, first of all, as I said, I think AI will make the strategic end of procurement more human, not less, because we will be automating the mechanics. So the premium will go up on commercial document, on judgment, on influence, and on strategic supplier relationships, because that part is not going anywhere. And then the second part I think that the scarce skill that will be in demand would be the ability to define what good should look like because you cannot automate something that you haven't articulated. Yeah, it could be the old threshold, the specification, you name it. So in my view, people who can articulate their judgment will become much more in demand than people that use their intuition and just their intuition alone. This is the second. And the third one, in my view, procurement will move more upstream and become more connected to their operating system of the business. Because when I look at my own industry, we're seeing an unprecedented supply chain disruptions. We're seeing lead times that are measured in years. Yeah. So you cannot really expect and wait for the project to be fully backed in and then come to supply chain to source it. So the teams, and this is their use case that I'm really looking for, the teams that can connect for example, the maintenance planning, the inventory, the supplier intelligence, the project demand, and connect it all together so they can shape decisions earlier so they will be ahead of the curve and they will be, yeah, they will be ahead of the curve and this will be the exact competitive advantage they will be bringing to the company. So this three things. Can I just ask a follow-up, which I think kind of ties some of these together? We've often been, as we've grown up in procurement, grown up in supply chain, I think that it's easy for our mindset to be constrained around what we believe that procurement is and what we've been brought up to believe the role of procurement and what we've been brought up to believe the role of supply chain. How important is it to be for that talent when you think about who's in demand? To be able to see past those, break down those walls and start with a blank piece of paper and think, okay, now we actually have the ability to do far more than we used to. Where does procurement start? Where does supply chain start? Where does it stop? And how is that different from perhaps what we've all been used to in the past? That's extremely important, Phil, and I think you nailed that. The ability to not to apply our preconceived notion and automate potentially a bad process, but start from scratch, from their clean slate and then build something that is fit for business where the process is not dictating the activities, but you just look at where the company is going and how you can accomplish or help the company accomplish their goals using whatever tools that you have in your toolbox. And it could be AI, it could be, it could be clean data. And sometimes clean data will be AI. We just don't know because do you really need AI when you can serve? Yeah, it's difficult to tell right now, but I think we can anticipate a lot of changes in our industry and just as a whole and procurement. I think we become architects. You know, we now have all these tools at a disposal and the question is how do we fit together the right tools for the right for the right moment for the right stakeholder for the right project? And the answer is not what it used to be, which was, okay, I'm going to throw a sourcing process, you know, I'm running an RFP at that. It's, you know, how can I build something that feels like it's really custom for this moment in time, but obviously in a way that is still scalable. That's going to be hard and that requires a different mindset, I think. A different mindset, but the mindset should be start with an outcome in mind and then build everything else based on that as opposed to start with a process and try to fit in the outcome somehow. Well, Lisa, I want to thank you so much for joining me today on the podcast. It's been a pleasure. What I'll do is we'll link up to your LinkedIn profile in the show notes. So anyone who would like to reach out can find a way to get in touch. The show notes will be out of procurement.com/podcast. And one last time it's been a pleasure to speak to you today. Thank you, Phil. It's been a pleasure as well. If this episode struck a chord with you, please do send it to somebody. We grow here at out of procurement through word of mouth. And that would be really appreciated. You can also support us by giving us a thumbs up, a star rating, a review, wherever you listen to your podcasts. Since 2015, we've built the world's largest free resource for procurement professionals looking to elevate their impact. Our resources span podcasts like this, videos, blog posts, papers and events. To join us on the inside and to ensure you never miss an episode of webinar or events or a post, please do subscribe to our weekly newsletter this week in procurement. You can do that at artofprocurement.com/subscribe, that's artofprocurement.com/subscribe. Thank you for joining us today and I look forward to seeing you again soon.