In more than 12 years at PepsiCo Tariq Munir, has held roles including FP&A Manager, Head of Integrated Business Planning, Head of Finance Supply Chain, and more recently APAC Finance Transformation Lead (based in Australia) delivering Financial Planning process simplification. This included founding the first-ever APAC Finance Digital Academy to build a digital mindset and culture. Tariq is sought out as an international keynote speaker and is a regular columnist for CFO Magazine ANZ, sharing insights on digital trends, strategies for digital resilience.
In this episode:
• Running digital transformation at large companies
• The opportunities and challenges in your data
• Core problems faced by finance teams including transactions
• The big headache AI is causing for finance teams
• M&A and AI Transformation
• retrieval augmented generation (RAG) and plugging into external data source for your organization
• Framework and governance for digital transformation
Connect with Tariq Munir on LinkedIn: https://www.linkedin.com/in/tariq-munir/
Further reading:
Competing in the Age of AI: Strategy and Leadership When Algorithms and Networks Run the World
McKinsey: Gen AI: Opportunities in M&A
https://tariqmunir.me/
Full transcript
Glenn Hopper:
Today. Welcome to FP&A Today, I’m your host, Glenn Hopper. Today we’re joined by Tariq Uni, an international keynote speaker, and a columnist for CFO Magazine Australia and New Zealand. With over two decades of experience at multinational corporations, including PepsiCo, Axo Nobel, and PWC, Tariq has dedicated his career to transforming finance operations and building digital first cultures. He’s especially passionate about helping finance leaders navigate the path from traditional processes to digital transformation, not just through technology implementation, but through the critical cultural and mindset shifts required for success. He brings a unique perspective on building digital literacy and finance teams. Tariq, welcome to FP&A Today
Tariq Munir:
Thank you, Glenn. I’m so glad to be on the show today.
Glenn Hopper:
Yeah, I mean, I, I say it all the time. I love this show because I feel like I’m, uh, a lot of times I’m preaching to the choir and, uh, get an opportunity to speak to, uh, kindred spirits and people who are as passionate about this stuff as I am. And, uh, you know, just from the, uh, from our, our, our pre-recording conversation, I feel like, uh, I’ve, I’ve got a lot to learn from you today, so very honored to have you on.
Tariq Munir:
Same here, Glenn. Same here. Same here. I look forward to our conversations. I, I’m, I’m sure we are going to enjoy it a lot. And, uh, there’ll be a lot of takeaways for us, for both of us, and for the, for the audience as well. Of course.
Glenn Hopper:
So, I think you and I had a similar shift where, yes, we’re still finance people, but we’re like the finance tech stack people, and we’re focused on data and automation and a lot more than just traditional, you know, debits and credits and, and, uh, variance reports and all that. So talk to me about sort of your transition. What inspired it going from the more traditional finance roles to becoming this a, a digital transformation leader?
Tariq Munir:
So, Glenn, I’m an optimist, an optimist in technology, right? So I do firmly believe that technology, especially the general purpose technologies, we talk about, like ai, cloud computing has this potential to transform how we work, interact, create, and for good reasons, right? I’m very confident that these, these technologies are going to create, uh, a future of a prosperity and some of the most complex human and business problems that we face today. As a leader we need to create. We need to craft a future, which is both enabled by tech and humanistic in its center and its core as well. And I do believe that we need, uh, more business leaders focused on creating such a future as opposed to just doing the, doing the debit credits and, and, and stuff, of course, so we can shape up the future in a way that is both tech, tech enabled, of course. And that is also solving some of the most complex problems that we are facing. And, uh, to be honest, that’s, that’s what inspires me. That is what, wherever I have seen an opportunity, I have raised hand my hand, and it has put me or landed me in places and roles that involve digital transformation in one way or the other. So that’s something that drives me, that’s more of my mission as well, that that’s how I, I keep myself inspired to keep going on this journey. In other words.
Glenn Hopper:
And for me, it was a lot of it was necessity in, in that transition. And I found that gaining efficiencies come from, well, first off, if you, because I was in the startup space and I, you, you’ve, you’ve got the advantage of being at massive companies that have all this wonderful data. My, my initial, um, you know, my first CFO role was at a company that was doing under 10 million in, in revenue, and we just didn’t have a lot of data. So it was kind of a, a blank slate in, in trying to figure out how to report when, you know, a company’s on QuickBooks and GL is kind of the only data you have trying to build out that architecture. But when you’re trying to raise money and, uh, uh, trying to present as if you’re a bigger company, you’ve gotta be able to have some kind of data that you can, that you can report on.
So building out that was, was important, but also not a lot of resources. So you start thinking about software and automation and, well, is there something off the shelf we can use, or do we have to build something? I imagine it’s different being at a such a large company running digital transformation versus being at a startup because it’s like, you know, startup being the speedboat that you can quickly turn. But these big companies, it’s like turning a battleship, you know, it’s, it’s a massive thing to, to transform and change the company. Tell me about like, doing digital transformation at a large company. What are some challenges that you face there that you may not in the, um, in, in smaller companies that are able to be maybe a little more nimble?
Tariq Munir:
You are absolutely right, Glenn, and, and you pointed it out very well, that data, of course, there is a lot of data, but data has both sides, right? It has a, it is an opportunity if you look at it, but it’s a challenge as well when you look at it the other way. So again, a lot of, in based on my experience, what I have seen that, and in many roles that I have done, a lot of insights actually sits, reside somewhere in, they are there, but they reside somewhere in, in complex Excel models, in, in disparate systems, which are not talking to each other, or things like reconciliations and so on, so forth. So I have seen a lot of work or, or a lot of focus being done on carrying out, just reconciling and doing all these work, which is more of a manual core accounting piece, as opposed to then creating value for business, which is more around business partnership and those areas.
And, and there’s nothing, I’m, I’m, I’m, I’m not saying that it’s, it’s, it’s right or wrong in a way. It’s just that the nature of many of the finance functions that I see today, that is how, how finance functions are, are unfortunately operating. So data is a big, big challenge, I would say that, that one, one faces, and then of course, when in, in a bigger organization, you have metrics structures, you have a different way of navigating digital transformation as opposed to maybe in a smaller organization where you have a much nimble approach towards decision making and moving, moving faster. So in a bigger organization, you need to have a much more coherent approach, a much more clear vision of where we, you are trying to go, right? How do you actually get there? Having a clear roadmap. So all those things are very, become even more important in bigger organizations, uh, in a smaller organization. Of course, as you, as you rightly pointed out, it’s, it’s nimble. You can, you can move fast. You, you, you don’t have to, uh, you know, think about a lot of different areas of the organization because it’s small in, its in, its in, its, uh, in its very nature as well.
Glenn Hopper:
Yeah. You said data has both sides, and one thing is, but at the big organization, the one advantage you do have is more data and data. We talked a lot before the show about ai, and everybody’s talking about AI with generative ai, but you and I were talking more classical AI and, and you said, you know, you called yourself a traditionalist, which I would completely agree with that, and I think, and we’ll talk about it more, but I think that something that people have a hard time with is classical ai, machine learning is a deterministic science. Mm-Hmm. <affirmative> versus generative ai. It’s <laugh>. You could get a different answer each time. And, you know, there’s obviously guardrails in, in ways that you can, can use it, uh, for finance. But let’s back up and talk to me about what you mean when you say you’re a traditionalist when it comes to AI and finance. Like, explain, you know, why you believe finance departments need to focus on traditional machine learning before they jump into the generative AI side.
Tariq Munir:
Yeah, great, great question, Glen. First of all, let me clarify one thing. I have no doubt on the potential of generative ai, right? Being, being all, all in for technology, and you know, how I meant how I talked about how technology can revolutionize, uh, what we do. So no doubt about that. But having said that, as we already mentioned, that many organizations today, finance, in fact, some of the studies also says, suggest that around 40 to 50% of the time, finance is still spending on transactions, right? Mostly transaction processing or, or the other problem around inefficient processes, right? Again, I mean, that’s one of the, one of the challenges many finance functions face over the period of time, m and as acquisitions, and, and just by organic growth itself, things have become so complex that, um, we waste a lot of time in those areas.
Now, what generative AI has done, unfortunately, i I say, is that it has diverted the leaders, business leaders and finance leaders attention from these core problems, right? What it has done that people are starting to see this, this sexier interactive AI experience, they, they, they, they have started to have that experience where you can actually talk with ai, it responds, you see the result right away, instead of a traditional AI where you need to do all those, all those boring stuff, and you don’t really see an output in a way that you actually see in a generative ai, which is very, very interactive and very on the spot, on the spot output, on the spot result.
Now, it’s a dangerous trajectory. I say the reason being a technology with a lot of potential, no doubt about that. When we put disproportionate resource and attention on generative ai, which has less potential at this stage for the traditional processes like transaction processing, process improvement, those, uh, kind of reconciliations and stuff, when we put disproportionate amount, we see the organizational technical debts keep on increasing. And then we see that, you know, I mean, most of the projects getting into the, what we call a pilot purgatory, right? So, so, so that’s, that’s a dangerous trajectory to be on. What unfortunately, generative AI has done is move that shift, move that attention. I hope it, it, it makes sense to put some of the context around why I feel that generative AI is probably doing a little bit more harm than, than, than, than some of the benefits that we are actually able to reap today.
Glenn Hopper:
Yeah. And actually, maybe now it’s a good point. So for finance professionals who might be new to the space, I mean, again, we are finance and accounting people, not machine learning engineers, <laugh>. Yeah. So explain the difference between, uh, the differences in their respective applications of, uh, classical AI machine learning versus new generative AI.
Tariq Munir:
No, absolutely. So as, as you mentioned as well, that traditional AI or, or machine learning, what we say is more deterministic in nature, right? It learns based on, based on historical data, and then it predicts output based on a very specific outcome. It can, it can predict results, it can classify information, or it can detect anomalies, for example, right? It is, of course, powerful at, at, at pattern recognition because it is an ai, it is a machine at the end of the day, and it can, can carry out complex analytics as well. So, for instance, if for in a traditional view, in a machine learning, if you want to predict your future cash flows or identify a fraudulent transaction or segment your customer based on purchasing patterns, you need to write down that algorithm. And that is the output it’ll give you for a specific cash flow forecast.
It can tell you, okay, this is how your next three months look like from a cash flow point of view, for instance, right? Generate AI way on, on the other hand, as the name suggests, it can generate new content, right? So for instance, predictive analytics can give us a cashflow for the next quarter, but then generative AI can then go a step further and can provide recommendations based on them. Of course, all those, all those language models, all large data that it has been trained on, on the, and the large language models, it’s, it’s based outta, and then it can also provide, like, provide, uh, summaries or variance analysis and, and anything that can generate a new content. Now, having said that, today, most of the technologies, or the underlying architecture that sits under traditional AI is also part of the generative ai, right? So a generative AI effectively can do predictive analytics in, in, in a way.
Uh, and then, uh, the, the, the, the layer, which actually works around this generative part, can sit on top of it with whom A CFO or, or, or a team can interact to actually interact with those numbers. So it, it pretty much adds a layer. If we step back a little bit, the core still remains that traditional AI, the core of finance transformation or business transformation still sits with that customer segmentation, cashflow forecasting, process mining, for example, process optimization and so on, so forth. Generat AI on top of it at this stage is more of how do you actually interact with those results and then come up with new ideas. Maybe
Glenn Hopper:
Right now it feels very, you know, you can do a lot of parlor tricks with generative ai. You know, it’s, it’s like a magic trick, and you can impress people and you can do some, uh, pretty amazing things, but it’s not gonna be fully mainstream. And don’t, don’t get me wrong, I’m all in on give your teams access to a generative AI tool. Make sure it’s in a, an enclosed environment where it’s not, you know, the data’s not gonna Yeah. Be leaked out and, and used to train models and all that. I think that it’s generative AI in finance and accounting isn’t gonna be wholesale widely adopted until it’s integrated into the software that we already use. Mm-Hmm. So data rails, you know, has, has AI built in where you can ask it questions. Yeah. And I think you’re gonna see ERPs, CRMs, the, you know, Salesforce obviously huge on, uh, on generative ai, and as our, as this oracle with, with NetSuite, and the list goes on, but I think you’re gonna see those software engineers are gonna put it into their systems you’re already using.
So if you’re used to seeing a dashboard that is, you know, giving you two dimensional view where you can, you know, dashboards, you can click in and, and see additional data, what you’re gonna start seeing though, is the ability to get even more detail in natural language. So you’re not having to write a SQL query. You’re not having to know Python to be able to slice and dice the data. It’s gonna be integrated there. And if, if companies have their own, you know, you know, machine learning, data science, uh, BI teams, they’re gonna build applications internally. If they don’t, though, they’re gonna be, they’re gonna have to wait for these software providers for it to come out.
Tariq Munir:
Absolutely. Yeah. Yeah. So, so true. So true, Glen. That’s, that’s the thing, right? I mean, what I do firmly believe as well is the convergence of the technologies, right? And that’s where the real benefit comes into play, right? When we actually start seeing different kind of ERPs and, and traditional AI and generative AI all converging into one, evolving, just like human evolution, like AI is like an inorganic bean. And it is, it is evolving, however, I do call it that. It’s, it’s from a, currently from a traditional ai, traditional AI is a single cell <inaudible>, then the generative AI is nothing but a multicellular organism still far away from becoming a human <laugh>, as we tend to believe, uh, or we, we actually think like, you know, it’s, it’s so advanced that it can solve, pretty much, solve all our problems. It’s more about taking a reality check that yes, it is, it is, it’s, it’s cool, but not so cool to solve everything that we have today. Yeah.
Glenn Hopper:
And that’s why I always push for, you know, if you’re gonna be using these tools, especially if you’re gonna be kind of out on the leading edge using these tools, you have to understand how they work. Not saying you have to be a, an engineer, but if you’re gonna be using your, these tools to analyze data and all that, you have to understand roughly how it’s working. Otherwise, you, you know, you, you, you, you, you can’t, you can’t tell the auditor, you know how, oh, how did I get this number? I put it in the magic black box. Yeah. And this is what came back. I mean, that, that’s not gonna fly, but it is. I also, because I talk about this a lot, I do get opposition from, mostly it’s the more senior, uh, seasoned, uh, finance leaders who say, I spent my career learning to be the best, uh, you know, finance and accounting person. I can, I’m, I’m not a com a computer developer for, if I wanted to do that, I would’ve gotten my <laugh> a master’s in computer science, not in, uh, in finance or whatever.
Fp and a today is brought to you by Data Rails. The world’s number one fp and a solution Data rails is the artificial intelligence powered financial planning and analysis platform built for Excel users. That’s right. You can stay in Excel, but instead of facing hell for every budget month end close or forecast, you can enjoy a paradise of data consolidation, advanced visualization reporting and AI capabilities, plus game changing insights, giving you instant answers and your story created in seconds. Find out why more than a thousand finance teams use data rails to uncover their company’s real story. Don’t replace Excel, embrace Excel, learn more@datarails.com.
I know with your digital transformation work, you’re very passionate about upskilling finance leaders, and I’m wondering if you see that opposition, and if so, how you overcome it? What are key challenges to build that sort of, I call it digital literacy and, and get that mindset among finance people to speak this language and use this technology? Like, how do you help finance professionals kind of evolve themselves and transform themselves to be able to understand and work with all this?
Tariq Munir:
A lot of this apprehension towards digital literacy or this protected tough kind of a thing that, you know, yes, this is my turf, my turfis accounting. I need to focus on that one techno, this is technology. I don’t need to know about it. I have IT have stuff. A lot of that is, is coming out because of number one, a fear of unknown, right? People don’t know. We don’t un it’s, it’s an, it’s a, it’s a lack of understanding. I would say that how much, and you rightly pointed out as well, that how much digital literacy is needed, actually, do I need to know how to code? Because if, for example, if you’re asking a CFO who has been in the role for 20 years, or, or who has 30 years experience, I mean, you will talk about digital literacy and you, the first thing that will pop up will be probably to her will be that, you know, I need to start doing the coding.
Right? She might not be interested in that <laugh>. Okay. So I believe, first of all, it’s important, and that’s what I, I mean, I, I did some of the, um, I led a couple of initiatives around building a digital literacy and a digital mindset, uh, within the organization, within the finance team, specifically a big APAC finance team. And I believe the first and foremost thing was to educate people on in understanding that what is actually a digital literacy all about for a finance person purview, what it is all about. And as you also also mentioned, it is not about coding. It is simply understanding. As a rule of thumb, it is simply understanding that how the output of an algorithm or machine can help you solve your specific business problems. That is it, it is all about just understanding that it is about understanding how machines are calculating and arriving at those insights, how data is generated, categorized, stored, and then consumed within the organization.
That is all we need to know as a finance leaders. Once you build that clarity that this is what we talk about when we talk about digital literacy, it opens people up. It makes them warm towards, yes, this is something I can learn, right? This is something I would like to know. Now the next hurdle comes. Once you do that, the next hurdle comes, you have an over. We are fortunate and both unfortunate to be living in a world which has an overload of information. It is very hard today to go out. And if you want to understand something, if a finance person wants to know what to do with, uh, for example, what to learn about AI, there will be thousands of training courses available out there. So I do believe that where we need to focus, and that is what I personally have done as well, that instead of, we, we, we have a lot of content already on the, on, on the internet and everywhere.
So creation of content is not an issue, and quite a good quality content available everywhere. It’s all about curating the right content for the right people. So that is where I do believe we need to put our focus more, and that is where I have put our focus. And that has worked, actually, people appreciate that. Things bringing out in front of them, which are very clear pathways, which are clear, clear roadmap towards, towards their upskilling journey. Yes, you need to put a lot of work and a planning behind that, but once you do that, you can build that culture. One thing I would, however, like to mention here as well, and that’s to your, to your initial point also around that leadership piece as well, is about building a continuous learning culture within the organization, right? That’s, that’s the key. Now, one misconception, and, and I’m also culprit of that, right?
I wouldn’t say that, that’s just, that’s just, uh, uh, others. I’m also included a misconception around being continuous learning, is that the more hours we clock or the more number of certificates we earn in a, in a month or in a year, that means we are becoming a continuous learner. Unfortunately, that’s not true. Continuous learning is all about how much we are willing to update our knowledge and our assumptions in the light of new knowledge. So when we are exposed to something new, are we ready to stick to what we know or we are ready to open up ourselves and update our assumptions? And that is where the leadership role comes into play. So leaders need to lead from the front and demonstrate that continuous learning behavior that yes, I don’t know everything, and I’m okay to go back and learn more. And when you, when you say that, yes, I’m open to learning, your team will say, I’m open to learning as well. And that’s how you create a digital culture. That how you create a continuous learning culture or a digital mindset at the organization
Glenn Hopper:
That’s very of you. The, uh, idea of <laugh> updating your thoughts based on the new knowledge as it, as it’s presented. It’s almost cliche now. People say it all the time of, oh, you know, business is moving faster than ever. And, um, you know, I feel like we’ve been saying that for 30 years. We’ve been saying digital transformation for 30 years. And what I’ve found a few years ago is when I would say digital transformation to people, they would say, I feel like we already digitally transformed. I feel like we did this in the late nineties, <laugh>. And I think it’s because if you say transformation, people think of it as one and done already transformed. But it’s, it’s kind of like it goes hand in hand with the continuous learning. True CPAs obviously have their, their, uh, CPE credits they have to keep learning on. If you’re just a finance person, as as I am, there’s no professional mandate that you keep learning. But if you don’t keep learning in finance, you’re gonna be, I mean, whether it’s the latest software that’s available, the latest methodologies that are a available, you’re gonna get left behind. So it’s interesting how digital transformation and continuous learning go hand in hand.
Tariq Munir:
Oh, absolutely. Absolutely. It does. They have to go hand in hand, because otherwise, because see, at the end of the day, what happened in nineties, and I, I totally agree, and, and I can totally resonate with that, most of that, that was the first layer, first wave of digital transformation, I would say, which was mostly centered around ERPs and mostly around accounting systems, reporting systems, consolidation, and so on, so forth, which was very digital in nature. There was a transformation element, but it was very digital in nature. It was a one set digital deterministic piece of technology you put in. You, you go, it goes live. And it performs in a certain way, AI or cloud computing. These are entirely different ballgame. You have to continuously evolve yourself. So I always say that digital transformation is more about transformation than actually digital. It’s just that we have to put a word to it, right? So <laugh> Yeah. To make it to it so that it makes sense.
Glenn Hopper:
And I do wanna get, we’ll, we’ll, in a few minutes we’ll get to the, uh, sort of the non-tech part of digital transformation, but I do, um, a couple more questions on, um, generative ai, just because, you know, I love when we have a, a guest of your caliber on, and we can talk about this, um, before the show, we were talking about that McKenzie case study about using generative AI and the M&A process. And I think as someone who’s done a, a lot of transactions over my career and thinking about the long days and late nights and putting together data rooms back in the days when a data room was actually a physical room <laugh>, um, room. But tell me about that case study and K kind of walk us through how AI is transforming the m and a landscape, everything from target identification to, uh, post-merger integration.
Tariq Munir:
That’s a very interesting case study actually, I came across. And, uh, as you pointed out, that in, in M&A you, you first step is probably to find the potential targets, right? To, to to know what, what you are looking for, which industries to target, which specific stage an organization is, whether it’s a startup or it’s in a growth phase or, or, or, or a more mature organization. So McKinsey published this study, a couple of, uh, a couple of, some some time ago. They have a proprietary tool which scour through and, uh, sift through a vast amount of data of different repositories out there, and it scours through thousands and thousands of potential targets. And then based on the, uh, specific acquisition criteria, it then creates a list or a more of a visualization of, uh, if you, if you actually search out the m and a McKinsey, uh, it has a very good, a good, uh, uh, visualization of how it actually tags all those different targets based on their different characteristics or based different criteria.
And then once you do that, uh, you can actually then, uh, shortlist those based on certain prompts on what are those, out of those safe 1000 or 2000 potential targets. You bring it down to 30, 40, 20, whatever is your specific criteria. Now that’s, that’s, that’s one part of it, right? On the other hand, where it can, it is actually creating a lot of benefit and productivity is around the due diligence process, summarizing your key documents. Like, I mean, you can, you can imagine in m and a how many documents you have to sift through, right? What are the potential risks? You don’t have to have a highlighter to highlight each and every, every, uh, every contract, right? You can draft initial memorandums, you can research relevant legal and regulatory information, which is sitting probably outside of your, your internal, internal databases as well. And generative way, I can can, can do that.
There are techniques there. Now, another way they are actually helping, uh, uh, helping their, their organization is around analyzing a lot of past m and a performance. And then based on that, they are able to generate personalized training programs for the new people, for the onboarding, for the onboarding of the employees. And then it also helps them with the integration piece, right? Because again, at the end of the day, once everything happens, you have to integrate the new, new organization into it. And it can, it can actually, based on the different m and a playbooks that the organization have, or the different, uh, policies, procedures, and frameworks have, it can actually bring out that, okay, these are the different, these are the nuances. This is how different organizational structure, the new organization, this is the structure of the organization. This is the structure of your organization, how we can bring it in line.
It can recommend you that, uh, some of those integration advice. It can give you an advice act as more of an assistant assistant for you on that one. I do believe it can revolutionize the m and a space. And, uh, it can, it can create a lot of productivity in that space. It makes our lives more humanistic. I know how much time m and a & a people spend in, in actually doing, going through a transaction when it’s a transaction is happening, that’s their life. Uh, but I do believe that again, you know, AI has that potential to make it more humanistic for people as well, so that they focus on the areas which are more important, which are more human, which are more value added, as opposed to just sifting through the documentation and everything
Glenn Hopper:
During the time of due diligence and going through an entire transaction. You still have to do your normal, um, finance and accounting task. You still have to close every period, and you have to do your reporting and your forecast, you know, let alone what’s gonna happen, uh, post merger. And I think about, you know, the, the expectation that we do all that. And then, you know, in the, in recent years, ESG been thrown on top of us. Hmm. Across the board, talking to CFOs. You think OO okay, this is in the office of the CFO, how, I mean, and I guess because it’s tied to compliance, which goes into the CFO, but now you have this whole new level of work that has to be done. And I know another area where generative ai, um, there’s, you know, we could potentially, uh, see some efficiency gains would be in the ESG reporting. So how do you see generative AI helping finance, finance leaders address ESG reporting and sort of mitigate kind of the greenwashing risks and, and, and everything around ESG?
Tariq Munir:
You’re absolutely right. Uh, Glen, I, this is, this is a new role coming up, right? And the, the, from an organizational placement point of view, CFOs are the ones who are actually most, in most of the organizations, CFOs are the ones who are eventually responsible for the reporting piece, the compliance piece, to ensure that there is no greenwashing happen happening. And we have seen, uh, lately, even in Australia, we saw some of the fines through the regulator on, on some of the organizations going through that, uh, that greenwashing. Now, intuitively, when we think about generative ai, I mean, similar to how we talked about m and a, that is how it, it, it works, right? I mean, intuitively you can train it on the data and it can then help you with all those, uh, compliance related stuff and audit. But it is good to understand what the way I see it’ll work out from a more technical perspective as well.
So, for example, there is this technique called, uh, retrieval augmented generation, right? Or, or rag, uh, you would of course, uh, know about it. So where a large language model is trained on a company specific static data, your policies, your procedures, but we know that in ESG space, a lot is happening outside of the company domain, right? There is so much of regulation, regulatory updates. And then there will be, uh, publishing from other similar organizations in a public listed space, their reports, their data. You are doing scope three emissions, which are, you are not really able to control. So how do you bring that up to date information into your large language model, which is sitting residing within your organization? And that’s where this, this rag technique comes into play. And without going into a lot of technicalities or details, it’s pretty much plugging into, for, for the audience, it’s more about plugging into an external data source, which is, which always stay up to date.
And then you can bring that, that external data source within your large language model, which is being used for, uh, uh, for your internal reporting and stuff through your internal model. You can do all the reporting and analyzing and summarizing of the stuff, but then through the external data, you can bring in, bring in sources like corporate reports, regulatory filings, industry standards, which are changing, or any regulations which are being passed on. So that can create a, a holistic view of how do you take your ESG, not just from a reporting perspective, but from a compliance and a strategy perspective as well. And that is where I believe the future is gonna gonna move towards. And I know there are solutions out there which are actually working towards that space of building rack capability within the local law language models. And, uh, take it to that, take it to that level where you can actually, uh, build that visibility.
And, and what it does is also that, again, you know, in, in, in, in, for not just for finance, for everyone, it is important. Accuracy is important. We have this concern around hallucination and around incorrect out outputs. And large language models are designed to give you an output. And one way or the other, they will just give you an output, whether it’s correct or not. Rags can actually help bring that, that piece of sanity within that large language model. And that’s very important from an ESG point of view, from a greenwashing point of view, from being, because you are now, more and more organizations aren’t, it is now become becoming part of your public, uh, statements that you’re issuing, right? So it’s your, it’s your part of your earnings call, for example, now. So it’s as important as, as any investor making a decision based on your quarterly earning. Now investors are making decisions based on how you are reporting or what are you doing on your ESG front. So that’s why it becomes very critical and very important to be a hundred percent accurate and not just rely on probably element and, and maybe look at different ways of doing it. And drag is, I suppose, is one of those, one of those ways it can actually transform.
Glenn Hopper:
Yeah. And that’s really where it’s gonna start making a lot more sense is when you’re using RAG and you’re using, uh, AI in your workflow and where you have live data coming into it. So, uh, you know, updates to accounting regulations, updates to policies and all that where it’s just, where it’s in the workflow where you don’t have to kind of stop and step out and go use this bot over here that’s querying our, our vector database of the, these rag files and everything where it’s just all seamless and I think it’s coming sooner than
Tariq Munir:
Later. Yeah. And I think one very important aspect is also that whenever a regulation change or whenever something major happens outside of your static data, you don’t have to retrain your entire line language model,
Glenn Hopper:
Right?
Tariq Munir:
That’s, that’s a cost in itself as right. It’s not a small and easy thing to do. So, so, so you don’t have to do that, and you are always plugged into that. And it’s a very fascinating technology. It’s a very fast, fascinating technique, uh, to, to bring that in. And, and you are right. I mean, it’ll help us with that workflow and, and, and, and so on.
Glenn Hopper:
Yeah. Going from that to really thinking about integrating. So I know you have a background in supply chain finance and FP&A and reporting, and I think about, and, and maybe you’ve got some examples, but if you can think about some of the, kinda the biggest data and process challenges that you encountered in knowing what you know about generative AI and kind of seeing the, the path there. How could you see emerging AI technologies addressing those issues that, you know, years ago would’ve been, that gave you heartburn years ago, but now maybe there’s a shining star at the end, end of this future where, where we see potential solutions that technology will bring for us,
Tariq Munir:
I guess data, we, we, we saw data was one of the problems. And of course, I mean, we, we can, we can see today, we can find today that AI is actually helping a lot with harmonizing the data and, you know, creating those workflows of data. And that’s where I generally mostly talk about is that at the end of the day, transformation is all about a mindset shift, right? It is, it is, yes, it is a lot about technology. And one key mindset shift that we need to make is around process excellence. And that is where finance or any business function, to be honest for that matter, is, is, is struggling, right? Our processes are so complex. There are over the period of time they have become, become so integrated or disintegrated with each other that it becomes very difficult, uh, or have using a technology might not help, uh, it will actually, as Bill get said, we would actually probably augment or, or, or amplify its inefficiency as opposed to eliminating that.
I always talk about this, uh, very exciting technology, not, not a generative one, but of course generative can always become, see, remember, generative can always be a layer on top of it, right? So at, at the core, uh, what I believe that technologies like, uh, tools like process mining and task mining, they are some real game changers that can actually help address some of those biggest challenges. So I had this, this experience working on, on, on, on a process mining, for example. That’s just fascinating to see that how within a system, how many touch points a certain document has, where does it go from where, and what’s the workflow you, it’s, it’s like a, it’s like a revelation that comes to you, and then based on that, you can then simplify your processes. And once you are able to simplify your processes, putting technology on top of it becomes relatively much easier and you are able to actually extract more benefit out of that. So that’s my one, one learning I would say is, is, or, or one, one key area that I do believe that technology can help us shift. If I had that technology like maybe few years ago, I might have, I might have done a few things in a different way. <laugh>
Glenn Hopper:
We’re, we’re preaching to the choir here, but I love that you can separate the, the technology from the transformation, because to me, so many people think that that software is just gonna be a, a magic wand, or AI is gonna be a magic wand, and you just buy this one piece of software and it’s gonna fix all your problems. But really you have to back up and identify the processes and find where the bottlenecks are and where the roadblocks are there, because otherwise on what cause are you picking your software? You have to know what problem you’re trying to solve, then buy the software to solve it or build this, you know, build the solution to that end, you have a whole framework for how CFOs should approach leading digital transformation initiatives. And it’s more than just a technology. So can you expand on that a little more too, on, on that framework?
Tariq Munir:
I would love to, I would love to, Glen, and, and you are absolutely right, and I am a huge proponent of this point of view that digital transformation is technology can only take you, take you so far, if you are not solving the right business problems, if you are not identifying the right business problems, and you do not have the right mindset and behaviors within the organization, no technology can do anything for you. Whether it’s the most cutting edge generative AI models, or it is the traditional AI we’re talking about, or some simple workflow automation. The way I look at it, the way I, I I talk to CFOs around how to approach their digital transformation journey is a, is a much more holistic view. So that I’ll, and I’ll give you a very brief of that, that that framework of course, first of all is around having a very clear vision.
Now, majority of, and, and right now today I was reading this report, uh, which was published I think a couple of, couple of weeks ago or a couple of months ago. 80% of a projects are still failing, which is twice the rate of normal. IT projects, in fact, 84%. A vast majority of the reasons of the, of, of the top reasons were people not leadership, not understanding the capabilities of what AI can and cannot do for them trying to solve problems which are not actually AI problems or even at some point overestimating AI capabilities. So the first step I always recommend to CFOs or business leaders is around having a, crafting a very clear digital vision, understanding why, starting off with your why, why you need to transform, what is it that bigger picture, what does technology gotta do with all of that? And in, in other words, we need to paint a picture where if anyone in the organization looks at that vision, they care about that vision, that why as to they should, they should be bothered about transforming, it should be moving, it should be very specific.
And then comes the, the building, the roadmap part or building the action plan part. That’s where the key action as, as the name suggests, that’s where the key action happens. And unfortunately, that is where we generally start off in a reverse manner, where we start off with the technology and end up trying to solve a problem while we need to, uh, start off with the other way around, identify all your business problems, create a register of your business needs, move on, go onto the, onto the next step on mapping your AI capabilities with whatever, or whatever your technological capabilities are at the moment, what you need to build, and then identify some quick wins. Uh, again, quick wins. I always say that, and I’m, I, I repeat that pretty much in every, in every talk I say that quick wins are not the smallest projects in a smallest possible time.
They are the smallest project in the shortest possible time with the biggest impact. So they need to have an impact. They must be solving a business problem, otherwise they’re not a quick win. They are just, uh, they can be a momentum builder or something, but they’re not really a quick win. So go for those quick wins and then iterate, do it in three to six months. Three to six months, hydrations. The third element of, of, of, of the framework that I refer to is around engaging the organization. So all about communication. How do you, some of the tips I can give, I always give is around having some key digital ambassadors or key digital, um, evangelists within your organization who are, who are tech savvy, of course. And they are not necessarily part of the leadership. They should not actually be part of the leadership because they can actually help spread your word around transformation, building that culture amongst their peers.
So they, they work really well in that space. And even, you can go a step further, I have seen successful organizations using crowdsourcing, internal crowdsourcing to craft their digital vision or craft their digital jour, digital transformation journeys. This helps you help organization with owning or finance functions with owning that digital vision or owning that digital transformation. The fourth element I always refer to, and that is probably one of the most important one, is governance. Now, governance is not just about about, uh, having, covering the risk or privacy or security. Yes, that’s one element of it. Governance is also about how are we tracking for against all the projects where who is responsible? Who is that one person or one, or, or that group of people who are responsible for the digital, for leading the digital transformation. Many a times I have seen those failures happening because there is no one person who is responsible for, for delivering or or accountable for being delivering the projects.
There is no clear responsibility for example, right? Responsibility, accountability, and, and those roles are not defined. So that’s very important, understanding what your key KPIs are that you are trying to achieve, what is the measures Yeah. That you are trying to improve, for example, right? Whether it’s just your financial measures or it’s your non-financial measures, which includes your org health scores or your employee engagement and so on, so forth. And lastly, the last element of the framework that I always refer to, and I do call it as the base of everything, is about building a right to succeed. And when we say building a right to succeed, it means building a digital mindset. What we all, all talked about in the part, in the, in the previous, um, uh, discussion as well around building a digital mindset, having a continuous learning culture in the organization, and very importantly, having a psychological safety for people to come up with the ideas, to bring out new things, giving them permission to fail.
When we talk about failure and when we talk about finance, people suddenly start jumping on their seats that, oh, it’s finance. We are not, we don’t have a permission to fail. But it’s not about mindlessly doing or, or failing in your core work or your regulatory compliance and stuff. It is about giving yourself permission to experiment and acknowledging that yes, if you are, you are doing experiments. Some experiments will be success, some experiments will be failures. So having that psychological safety that people are able to take that risk, that calculated risk, is also very important. So yeah, this was the very high level of, of the, I hope I’m, I’m, I’m making, I’m making sense. It’s, it’s a lot of information, but I just wanted to bring it to, to one, um, one uh, piece within, within, within few minutes.
Glenn Hopper:
Yeah, and that was great. And I think there is one piece of that that is easy when you’re in project planning to underestimate how significant this is. You know, we talked about you doing power BI implementations, and when you’re an advanced analytics lead, and we talked about change management and builds building user adoption, and it’s, so, I mean, there are people like I I, I’ve had to deal with like saboteurs who are so upset about something in their job changing that they’re gonna try to derail the whole project. Are there any takeaways in, in your implementations where you learned how to best manage, um, change management around this and to help people get, you mentioned the quick wins too, and I think that that’s always a way and sort of having the ambassadors out there and, and, and not trying to do it as a, as a one person, like, you know, everybody get on board my train, here we go. Whereas you’re, you’re making converts as you go, but is there, I mean, anything you can say about, um, kind of change management and building user adoption, uh, on these new technologies?
Tariq Munir:
I think Glen, it, it all comes to the, comes to the basics. What I have learned over, over, over my last years of experience, what we generally tend to do, do, whenever we are looking at a project or when we we’re trying to solve something, we tend to, as a humans, we tend to superimpose our perception or project our perception or project what we are thinking, right? We make an observation, we draw a conclusion, and then draw a conclusion that as that business is facing a certain problem. And then that’s what we project across everywhere. Now, that might not be, and that’s the problem. That might not be the problem which your stakeholder is actually facing. So the first step, and that’s where I always, always, I cannot stress that enough, is anchoring everything into specific business problems. Now, change management of course, is a big subject.
There are so many techniques out there, but again, the basics remains the same. Understanding what your stakeholders want, creating a dashboard which is giving an information, or even if it has an element of that generative AI of interactivity, if it is not solving the business problem of your stakeholder, you will have that, that, uh, pushback from them. You can never bring people on board if you are not trying to solve their problem. And it’s, it’s basic human, human nature, or basic human like, right? If someone is trying to solve my problem, I would always try to side with that person that, you know, I, I would, I would think highly of that person that, yes, you are now you are trying to solve my problem. This is what I have. So spending time with your stakeholders is very important. As a finance team, we love to spend time, uh, at the backend and try to come up with different solutions and look at technologies as well as different things.
That’s great. But till the time we go out and talk to our stakeholders and understand specifically what do they need and all that they need might not be doable or might not be true. There might be some, and you, you, you sometime realize that the problem that they have is probably something which is more of a behavioral issue of certain people or some process issue and not really has to do with any dashboard, for example, or any technology for that matter. And many a times that has happened, I have done workshops, I have done full day, uh, full day working with, with workshops, with, with with business leaders and finance leaders. Many of the problems come out to be either A, either a misconception or two people did not just end up talking to each other for a very long period of time and realize that no, if they would have just talked to each other, the problem would have already been solved <laugh>. So to manage the change, I think the most important thing is to manage people expectation and understand their problem, where they are coming from, what is that, what is in it for them? What do they, what problem are we trying to solve for them to make their life easier? And that’s when, that’s how it sticks. And it has worked for me. I mean, it’s my personal experience, but that is, that has worked for me. I hope that that, that that helps.
Glenn Hopper:
Yeah. Yeah. No, that’s great advice. And it is always, it always surprises me because I always just think, alright, we’re all gonna get on board and, and be excited. But there’s, it seems inevitably there’s gonna be at least one person or a pocket of people who are, are digging in their heels and do not wanna change anything because this is the way I’ve always done it and this is the way that I wanna keep doing it, and then I don’t like change. So we’re getting really long here, but I do wanna squeeze one more question in before we get to kind of our, our personal fun questions, just because I know you are out there keeping up with the latest in technology. So just I’d, I’d love your take on, looking ahead three to five years from now, how do you see the role of finance professionals evolving as AI and automation in this technology is just changing so rapidly? What do you see happening to the role and in tradi kind of the traditional finance functions and in what we’re outsourcing to the robots versus what we’re doing?
Tariq Munir:
Absolutely. So, uh, Glen organizations of today, or finance functions of today are no longer just typical input. You process something and you create an output, or just for that matter, people and processes or, or technology. It’s a complex interplay of humans, algorithms, digital networks, on top of all those people process systems that we have been talking about. Now, what technology is doing is, is what Kareem Lahan and Marco Yian refers in their book, competing in the age of ai. It pushes labor towards the edge of the processes. And, and that’s a great book, actually. I always have that in my, uh, yeah,
Glenn Hopper:
I’m, I’m looking for it on my shelf right now. Behind me.
Tariq Munir:
I’ve got sign copy as a, as a reference. It’s, it’s a, it’s a great book, amazing book, <laugh> this, this point. Uh, and, and it’s, it’s, it’s so amazing. And when I apply that to finance, I think of it, yes, if we automate our core, which is the core is what, what is our core? Our core is our accounting transaction. That is what we tend to believe what, what finance is all about. And in a way, it’s a reality as well, right? The the traditional finance. So if we can automate that core and we build our digital operating model, which is digital to its core, finance professionals will move towards the edge of the finance, towards the edge of the process. And beyond that edge lies different functions like sales, marketing, category, operation, and so on, so forth. So effectively we’ll move into the right place where we should be, which is the people facing position, in other words.
So we’ll be moving from our core work, manual work into more business facing roles, which is, which lies towards the edge of those processes. And that is where we want to be.
And I do believe firmly that, believe it or not, AI will increase our human interaction as we will see more and more finance people moving into these meaningful roles, into these meaningful places towards the edge of the processes, it’ll make us more human because we will be able to interact more with our business partners as opposed to sitting in the backend and just trying to do transactions and stuff, or creating reports or building dashboards. We have this obsession for dashboards, building dashboards, right? We need to get <laugh> again, maybe a discussion for another, another day. I would like to like to point out a couple of things that will be needed for the future of these finance leaders to survive and thrive in this, in that, in that world where we are, where we are, we are in those business for business facing roles, we need to be emotionally intelligent.
We talk about technology and we talk about, okay, technology and nothing to do with emotions, but it’s very important. We need to have emotional intelligence, being aware of what our actions are doing to other people. We need to build digital mindset. We talked about we need to have a continuous learning mindset, and very importantly, we should be able to think critically. Technology will give us output. We should be able to evaluate what that output means in context of, of the business problem we are trying to solve or in context of the business partnership that we’re trying to do. So this is how I believe that the role of finance is, is, is evolving.
Glenn Hopper:
That’s great. Great insights and, and great book. We, we need to put a link to that book in the show notes, the link to the McKinsey study.
Tariq Munir:
Yes,
Glenn Hopper:
Absolutely. All that as well. Yeah. Getting to our boilerplate questions. I always love to, to see what kind of responses we get on these. So the first one is, what is something that most people don’t know about you something we couldn’t learn just by searching you on Google.
Tariq Munir:
Okay. It’s very hard to say that you can’tFind anything on Google about <laugh>, about someone today. <laugh> in one way or the other, you will be on Google, but Okay. I’ll tend to believe that it’s, it’s, it’s, that’s not the case. And that, and Google does not know about Everything about me. So one thing I love is I love doing oil portraits. Uh, so actually, in fact, all these paintings that you see at the back, they, they’re all all mine. And this is something I just enjoy as it creates a lot of mental space for me when I’m not talking about technology, when I am not, uh, reading into the thick subjects around ai, machine learning and retrieval, augmented generations and stuff, <laugh>.
Glenn Hopper:
Oh, that’s great though. That’s a perfect balance of Right and left brain too. It’s, it’s just <laugh> that way you get, you know, to, to fully remove from it. Those are, uh, I beautiful from what I can see. So if there’s, uh, you know, in our final questions always where people can learn more about you, do you have any of your artwork posted online that people could, could check out or, Uh, on my Instagram, there is some. There is some. Okay. <laugh>,
The last question that we ask everyone, and we haven’t talked really about Excel at all. We’ve been all in in technology, but, um, what is your favorite Excel function and why?
Tariq Munir:
Oh, Al always vlookup.
Glenn Hopper:
<laugh>. Yeah. <laugh>. So I love, so I just had this conversation with someone the other day. I’m a VLOOKUP guy too, and someone told me that’s how you show how old you are. Yeah. <laugh>, everybody else now, you know, is gonna do index, match, match or power query, whatever. Yeah. But I’m going back to the old days. I know the nineties,
Tariq Munir:
They know X lookup, X lookup, which is much more efficient. But again, uh, uh, if over my last 20 years of career, v lookup has saved me thousands and thousands of times. So I think I still still regard that as my go-to for now <laugh>.
Glenn Hopper:
I love it. I haven’t heard VLOOKUP in a while and I’m so excited because that when I was a guest on the show, that’s what I answered with too. And I really, when someone told me that the other day, it, it never occurred to me. But that is how you show you’ve been doing this a long time is when <laugh>, you’re dropping the V lookup. I think so, yeah. So ri great, great episode. I guess our, our last question just how can, um, how can our listeners, listeners connect with you and, and learn more about you?
Tariq Munir:
Of course. So, so, so best, uh, ways. Of course. Uh, on, on LinkedIn, I’m very active on LinkedIn, so you can always reach out, DM me there or otherwise, uh, my email address is pretty simple. Hello at tarik muni me. Or you can also go on my website, which is, https://tariqmunir.me/ But I guess, uh, for a quicker response, generally it’s easier on LinkedIn or, or just DM me and I’m, I’m, I, I always respond to people. I’m, I’m, I’m generally good at engaging with, with, with the, with the like-minded people out there.
Glenn Hopper:
Well, thank you so much for coming on. Great episode. Loved your insights and experience. And I love your oil paintings too. That’s so Learn something new every day. <laugh>.
Tariq Munir:
No, I, same here, Glenn. It was, it was so, uh, it was a revelatory, uh, um, there were a lot of revelatory moments for me and I, I learned a lot as well from our conversation. I really enjoyed it. So yeah, looking forward to it coming online soon. <laugh>,
Glenn Hopper:
Thank you.
Tariq Munir:
Thank you so much.
