Navigating Major Programmes

RADHIKA: What we're actually seeing is that maybe at the point in time when contracts are signed, the right sort of steps are taken. But then whose responsibility is it to monitor? During the life of the asset, the right kind of investments are done to ensure that the environment or the community are protected and continue to be protected.
Evgenia: You're listening to Navigating Major Programs, home to the conversations Elevating the Infrastructure. And I'm Evgenia Angelina.
SHORMILA: I'm Shormila Chatterjee.
RICCARDO: I'm Riccardo Cosentino. And together we press the industry expert about the complexity of program management, emerging digital trends, and the critical leadership skills required to approach these multi billion dollar projects.
SHORMILA: Let's see where the conversation takes us.
Evgenia: Hello everybody and welcome to another episode of Navigating Major Programs. Today my guest is Radhika Anand. Welcome to Navigating Major Programs. Radhika, how are you today?
RADHIKA: Very well, thank you, Evgenia. Thanks for having me.
Evgenia: It's great to have you. I am really looking forward to this conversation. We recently met virtually. Your name came up, uh, through a number of substack articles that you've started publishing, which piqued my interest. Hence we're discussing and talking and meeting today. Uh, so lovely to have you. Um, before we get started, I'd love to ask you to maybe introduce yourself and tell us how you got to where you are and kind of what piqued your interest how you ended up in this big infrastructure space.
RADHIKA: Thank you so much, Evgenia. Well, uh, it's been an, uh, 18 to 20 year long career, so there's a lot to talk about. But, uh, I grew up in India. I studied in India and the uk. I started my career as an economist with Standard and Poor's. I helped build up their macroeconomic research capability for the Asia Pacific region. And at the time, um, I was analyzing countries across the region and I was looking at the fiscal policy, the monetary policy, the industrial policy, infrastructure needs, all of that stuff. So I was translating that for business decisions for S and P as well as for ratings, decisions across industry and um, different sectors and sovereign ratings. So I thought that, I think that work, the ratings work itself sort of formed the foundation for the way that I looked at the world, looked at, uh, whether it's the infrastructure world or public policy or economics, um, and I really enjoyed the work that I did. But over time, something was missing for me. You know, when you grow up in a developing country, you quickly understand that infrastructure is a, it's a means to provide access, it's a means to Provide opportunity and dignity in the society. Through the work that I was doing, through the education that I had. I had a top down perspective but I really wanted to understand uh, what really happened on the ground. So I think I wanted to understand the practitioner side, uh, more how investors and banks make decisions, how governments, you know, decide on projects, how risk get allocated across projects, how those choices ultimately affect um, outcomes. So I decided to go back to school and I did a second master's program at the Fletcher School at Tufts University. And I also took on courses at Harvard Kennedy in business schools. I already had a finance background, uh, so I went deeper into infrastructure finance, uh, regulation, P3s, um, investment law, contract law, that sort of thing. And from there on I moved into the infrastructure space and eventually worked across different sides of the infrastructure ecosystem. Over time I think the P3 space became particularly intellectually interesting for me because in all of the different elements of financing, public value, risk allocation, design, physical design of the asset, uh, operations, public sector incentives or private sector incentives, I'm sorry, all of that sort of came together. So this became an interesting, I think space for me. And um, I was quite fortunate to have been able to work across the different sides of the table and see the space, uh, and address issues from different perspectives.
Evgenia: I was just going to comment on the fascinating trajectory of your career and the combination of, you know, starting off as an economist and probably studying, you know, the, the greater the macro, macro scale, macro view on things, but then also how that relates to the social benefits. You probably have some personally conflicting views how that public good versus private good comes together from the socioeconomic point of view. Any thoughts on that? Did you struggle with that, uh, business versus social, uh, dilemma when it comes to public infrastructure?
RADHIKA: Not really. And thanks for actually asking me that question. It just takes me back um, a whole number of years. So when I finished my, my program at Fletcher, I was interviewing with the head of project, uh, finance at one of the banks in Asia, one of the leading banks in Asia, when the first, the first question he asked me is that do you believe in profits? And that's something that's really stayed with me till date because I think there is this misconception that a lot of people have, uh, that somehow profits and public value don't go together. They are opposite, they contradict or conflict with each other. I don't believe that. I don't have that conflict internally like you asked me. I think I've actually seen, um. I'll just give you an example of a very interesting uh, Experience that I had many years ago. And that's the reason why I sort of wanted to talk about the economics background as well. Because uh, so many years ago this was when I was still with Standard Poor's, I was helping uh, as part of some external engagements. I was representing S and P and I was evaluating Indian companies on their corporate uh, social responsibility and their environmental responsibility, how they were executing it, uh, for the purpose of deciding or shortlisting some companies for uh, uh, industrial awards. And uh, as part of that process I was able to go to rural India, very like remote parts of India, and see for myself how you know, when you make fiscal policies or monetary policies and you're sitting in sort of Delhi and Bombay or Toronto or Ottawa equivalent, you'll see that um, I mean many times at the grassroots level, what is happening at the center in the capital city, it doesn't really feel like it trickles down. Those people who are living there in the rural communities in these remote regions, even they are more impacted by the companies that are operating in uh, their communities. And again and again I saw the situation where many of the companies that I visited in their premises, they were very closely, and these are massive operations. So they were very closely engaged with the community. They were very closely. Not only were they working with the public administration where competencies and capabilities can be limited, but they were also working with them to develop public infrastructure, physical infrastructure. And obviously that fed into their you know, process, their production process, their, their supply chain. You know, it supported their business essentially their, in their entire manufacturing business and their sales and you know, supply chain management. But apart from that they also invested heavily in the social infrastructure of those communities. So building hospitals, identifying, you know, raising and uplifting um, the status of women in the society, uh, identifying uh, children from um, indigenous communities who could be really good athletes, the natural athletes, you know, naturally good at running or um, uh, any of the other major sports. There was a lot of work on uh, maternity care, postnatal pre. Yeah, a lot of work on, in those areas as well. Women's self employment groups, um, and of course, you know, they understood that if the people, the staff who's working for their companies is also living in the area, is also based in the area. So the water and the land resources that they're using is the same that their own staff is using. So anything that the company is very well aware and conscious even at that time that they had to use the land resources, the water very carefully, to the least extent possible for their processes. So they were Already aiming at net zero water at that time. And uh, I think they were also looking to you know, try and improvise the quality of life of their, of their staff. So long story short, I think this was a very insightful experience for me where I saw profit and public value being married really well, you know. And uh, so I think it's a matter of uh, it's not a question of whether public interests are opposite to or different from commercial interests. I think it's about finding the best way possible to marry them. It's a matter of your values as well as a society. What do you think is important? How do you cultivate that in the business that you do, in the infrastructure development that you do?
Evgenia: And I think what you're describing also I think with the developing countries a lot of the times those social benefits, in addition to economic benefits of obvious new piece of public infrastructure as hospital or um, community, community center or something like that, the social aspect of it, it really changes the uh, community around it. Like these projects have the power or so to do that. Do you think there is still, and we've known this and I think some of us who don't work on international projects or in developing world may not remember it or think about it as much. But in Canada, what are your thoughts? We obviously have programs in place around um, social benefits and community benefits. Do you think those programs are enough, um, in terms of kind of delivering that social value or we still can kind of learn from other examples while we're on the topic of social, economic, uh, value.
RADHIKA: Yeah, I think it's a good question. Overall I would say in across the world, uh, globally, I mean I'm still coming across case studies and situations where I feel like infrastructure could have been designed better. So instead of treating community or environmental aspects, uh, as add ons, I think we need to use them as inputs into the physical design of the asset as well as the financing structure, the governance, the commercial design of the infrastructure assets that we are looking to build. I don't think it would be, I would not be able to comment whether Canada wide, we're doing a great job of doing that right now or not. But I think generally speaking there is a lot that um, that can still be done in this regard. So we have. The good thing about developing markets and particularly in Canada I feel is that the institutional mechanisms are much stronger than what one would see in a developing country context. For instance, the ability to implement those mechanisms, the public sector, offices, regulators, the various sort of institutional players who are involved in the infrastructure development, they are reasonably empowered to enact and enforce their perspective on how they preserve the public value and the environment's uh, health as well. You wouldn't necessarily see that sort of thing across board in all of the, you know, especially the developing markets where I think there can be a bit of a gap there. But I will say that there is still a lot to be done. There is, you know, we've, we are trying to, for example on the indigenous communities, um, engagement side, we've, we've had the duty to consult for a while but you know, what does that really look like in practice? What does financial participation, which is a step up from duty to consult, what does that really look like in practice? Are we empowering uh, the members of the first nations enough so that they can meaningfully engage in the process is very important. Similarly, uh, when we talk about creating jobs or creating, buying um, services from service providers who are from the first nations, how are we designing around that? And is it a temporary, excuse me, is it a one side, is it a one time procurement, is it a continued procurement over the life of the asset? And similarly the same is the case for environment. Are we looking at environmental assessments as something that gives us information uh, at a point in time, or are we looking at something that there are things that we should monitor along the life of the asset? Uh, so there are some very, you know, interesting cases about this particular issue as well. You know, I was just reading uh, I was just reading about the water utilities in the uk. It's a much larger conversation and much larger debate, but they've essentially fallen to, fallen into the same situation where at some point water utilities were privatized. But uh, perhaps there wasn't enough investment done by the private companies during the life of the asses during their engagement period. And what has happened over time is that there are, there is sufficient evidence that there are, you know, now that climate change is happening, drought situation has materialized, uh, or drought risk rather has materialized. What we're seeing is that um, there's also contamination of water. So there's mixing of fresh water and you know, uh, sewage or non treated water toxicity, uh, levels have gone up. So I'm making a more general statement. I should like anybody who's closer to the subject would be able to speak better. But what we're actually seeing is that uh, maybe at the point in time when contracts are signed, the right sort of steps are taken. But then whose responsibility is it to monitor during the life of the asset the right kind of Investments are done to ensure that the environment or the community are protected and continue to be protected, uh, for that time.
Evgenia: Yeah, and it's a lot like with these long term projects and agreements, some of them the P3 framework, some of them elements of total privatization. We're still going through cycles of learning what the impact is and what the outcome is in terms of both, you know, the financial value versus the infrastructure itself as the asset class, how it withstood the time. And I think this is a good segue. So you've recently started writing more, uh, you have your own substack and you're sharing certain ideas and thoughts uh, around infrastructure and certain trade offs that we should be considering. Perhaps tell us a little bit more about that. You've made some provocative statements around, you know, the trade offs, uh, in infrastructure and how we're. Are they actually the trade offs or are we, you know, designing these failures kind of too early and then realizing that too late. You're drawing a very interesting, you know, not a conclusion, but kind of analysis. Um, and I think this is stemming from your background of having that economic business, but also public, private sector, commercial elements. Give us some overview of that thinking and kind of your worldview on that.
RADHIKA: Absolutely. Firstly, thank you for reading my articles. Thank you for taking the time. Well, I think for me personally I feel like I've come a full circle on the question that I started asking myself at the beginning of my journey. How infrastructure and business decisions get made. Uh, how could I make an impact in that? That was a question I asked myself years ago. And um, now I feel that over the years I've worked in advisory capacity, I worked with public sector, worked with private sector. I worked on the buy side a uh, little bit on the sell side. So I think it's been a very interesting uh, journey for me and it's allowed me to see whether it is infrastructure development or public service delivery decisions. I see that there are some patterns that keep emerging around commercial discipline. Again, like I said earlier, profit and public value are seen as though there are two things that compete with each other. Which I think is flawed. I mean, I don't think every single person thinks like that. I think many practitioners in the investment community, sorry, in the infrastructure community will agree with me. I think it's everybody's experience. But essentially what I'm saying is that what I saw firsthand was uh, how much value gets left on the table when the commercial architecture or the physical design doesn't fully design in what happens after the asset is ready what happens after the asset goes into service? You've prepared the asset, you've done all of the due diligence around it, you prepared it on time, on schedule, you've got your construction uh, sorted, you've managed the cost and everything. But what happens after that? So you have to design in the operational, community and environment perspective from the beginning so that you don't land up with surprises that are very difficult to solve for after the contract has been signed. By then you may not have enough flexibility or may not have enough options to fix things that were done wrong at the beginning. And whatever solution you might come up with is something like a patchwork because it's on top of uh, existing environment or conditions which don't quite align with what would have been ideal or would have been a smoother transition into something different. So I wanted to show through my writing, I've tried to include as many examples as I can without, uh, within confidentiality limitations and whatever is in the public domain as well, that if you do this with discipline, then you know, right amount of integrated planning, a long term view, then we have the ability to, you know, build a more fair or durable infrastructure asset. But when we are building assets in silos, it sort of locks people out and destroys capital. And that's the reason why I write because I want to bring that multiple, multidisciplinary or multi sort of view of. I think given my experience, I know how an investor thinks, I know how a banker thinks, I know how the government thinks, I know the challenges that government has in preparing project, in bringing them to the delivery stage and then how developers might approach the project as well. And of course the community and the environmental perspective in terms of um, I think your second question was around uh, design failures. So I've talked about, I've just said that many infrastructure trade offs are actually design failures. So essentially in my view, traditional approach that we have, that we take towards infrastructure development, it assumes that there are real trade offs, uh, into the process of infrastructure development. So investor versus investors, return versus public value, building fast versus equal participation, higher upfront cost versus predictability, development versus environment, and so on. So this debate is often along, it gets framed along the lines of investors and communities want opposite things. Now some trade offs are real, but in practice, like I said, how you design the physical asset, how you design the financing structure, the governance, the overall commercial architecture around the asset, that's what, what is actually critical to whether an asset succeeds. If the due diligence of commercial architecture before designing the deal are Weak, then the trade off is not inevitable. You've actually designed it because you haven't done the work that is needed. In the articles I've written four so far, I think two of them talk um, more extensively about AI, how AI is not a technology upgrade, it's an operating model change and how it needs to be approached. Especially when we think of generative AI and agentic AI. And then it goes on to the second sort of follow up article is about public procurement, uh, of AI capabilities. Uh, and then there are two other articles. One that focuses on what are the operating risks that need to be considered, um, at the time of making deals. So why should you think about designing a deal by thinking 25 years into the asset and backwards. And uh, also on data centers. So right now data centers are one of the fastest growing asset class. Federal governments, provincial governments are very keen to build data centers, but at the same time people are protesting and people are not very keen, uh, to have data centers in their communities. So what's the disconnect there? So those are the four articles and hopefully I'm trying to go down more as well.
Evgenia: That's super interesting. And you're touching on an idea that I struggle with sometimes and most of us are in this major infrastructure space, we pretend that uncertainty doesn't exist by putting in place certain commercial frameworks or certain design frameworks early on. But if I understand correctly what you're saying is that those early decisions are sometimes part of the uncertainty, that we can't grasp the level of uncertainty. A lot of the times when we start developing the project, be it AI, implication, technology, implication, environment, application, this is just the baseline, there is going to be certain level of uncertainty. What you're, I think trying to argue and correct me if I'm wrong, you're trying to apply, uh, that if we were to do a better job at taking the asset from the operating phase and kind of determining what the end result benefit to that asset is during operation, we should in theory do a better job at designing all of the contractual and construction, um, means and methods early on. Is that, did I understand that correctly?
RADHIKA: Yes, that's, that's really good. I think you've got the essence of it. So I can give you some examples. Uh, and I talked about them in greater detail in the articles that I've written. Um, but not to bore you and not to go into greater detail right now.
Evgenia: I love detail. Detail is good. Examples are great.
RADHIKA: Well, um, one example. So for example government, whether it's Canada or elsewhere, generally loves customizing technology or its services or assets. And they land up over customizing very often. Okay? And what happens as a result is that at that point in time it may seem that we need a very customized asset to meet our bespoke needs, which are very unique to us as an operator or a public service operator or an infrastructure operator. But then in the operations space you have to think 20, 25, 50 years depending on what the, what the service is. When you're thinking that far, uh, if you have an over customized technology, then scaling becomes very difficult and change management becomes very difficult because you have over customized technology. Any partner who needs to uh, be added to your system or any switch that uh, any capital development that needs to be done on top of your system might actually prove to be, you know, highly uh, risky to, very risky for business continuity or very, very costly or both. And so you're, you're forced to work with the incumbent. You're forced to expand and grow your technology or your service or your asset with the incumbent. So in the, in the articles I elaborated on, I think two examples. One is physical infrastructure and one is a technology company. So uh, this one is heavily sort of discussed in the public domain. The Auditor General's report, so referring to Presto, which is a major ticketing, uh, transit ticketing system in the GTH area. So when the card, um, when the fare system started, the contract with the incumbent was with the private vendor was Approximately I think $230 million or $232 million or thereabouts. And this is back in 2006. Now by 2020 the AG found that uh, it had grown by more than 1.7 billion. Now a lot of this I admit, was on account of the rolling in of the TTC and the Ottawa services, you know, folding them into the Presto services. So clearly there was a cost escalation on account of that, um, or rather contract expansion on account of that. But also the AG has repeatedly emphasized that it became, because the system, the initial technology that was developed for Presto was so heavily customized that subsequent capital developments became very difficult to procure from any other vendor, uh, competitively. So now you're locked into the original vendor as well as into the technology. It's not that easy to switch out of that technology or out of that vendor. And that's a pattern I've seen very often in the public sector. And it becomes a real challenge, um, when you want to switch out a vendor or when you want to procure, um, subsequent services, uh, in a competitive manner, you're limited by what you can do. So I think I've talked about more of such challenges.
Evgenia: What do you think is now that we've lived through a number of those experiences, and I think I can think of a few, um, the projects that I've worked and the clients I've talked to, is it the lack of, um, kind of commercial frameworks, like the customized commercial frameworks to use to kind of right off the bat to not have that problem where commercially you're limited to make a different choice, or is it poor understanding of your problem and the potential growth and kind of not designing for it? Uh, we know that the technology is the solution, but what's the question? What's the baseline? What do you think are we getting wrong in those specific examples?
RADHIKA: So I think many of these contracts were signed years ago. So hopefully there are lessons over the years like you said, and they are being applied. But I think the intention for me to write and write the substack is to actually bring them out as these are not siloed cases, but these are patterns. These are issues we want to standardize and fix, um, um, for every project that we look at. So hopefully some of this socialization of ideas and this discussion that we're having today can bring that into standard practice. Uh, I don't think that people are wholly, completely unaware of these things, but maybe they're happening in silos and they're not necessarily seeing all of these challenges in one place. So instead of thinking about what we're doing wrong, I like to think, how can we do things differently? Um, and these are choices. These are choices. So I've talked about. Let me talk about seven commercial risks that I've called out more specifically. So one, like I talked about customization, in many cases, these are like, um, almost like call options in financial theory. You can decide that there is going to be uncertainty around certain elements or decisions that you've taken at the deal stage or design stage rather. But you can plan for. You can create mechanisms that would allow you to preserve adaptability and flexibility during the operating stage without escalating costs or without getting stuck with one vendor, or without having issues with, you know, not being able to switch to another vendor or not being able to procure services competitively. So some of these issues are customization is a choice. Your particular service might require that you customize, uh, the technology or the services or the assets that you need. But know that scaling might be a problem, know that, you know, subsequent capital development might be a problem, know that interoperability might be a problem and build around it. Right. So I'm not saying customization is bad, I'm saying make that choice consciously. Similarly, what I've seen again and again, I've seen it more on the public procurement side. But also I think in the articles I've given an example on uh, physical infrastructure side. So the Ottawa Confederation uh, line that came up in 2019, and this is about integration among the multiple suppliers that you have. So for example, if an asset, the way that I look at an infrastructure asset, it's not a, uh, especially in today's time, it's not one supplier who's giving you the entire asset. You have a number of suppliers, number of service providers who need to work together to provide you the entire infrastructure, physical facility and the service around it. So when the Auto Work Confederation line came up in 2019, came to service, then very soon after it came into service, there was a problem with, I think doors are not opening or there were coordination schedule problems and so on. On paper they had contracted the integration capability with the primary sort of service provider rtg, but in reality that integration was not panning out. Right. And so we see this even in public procurement very often. What happens is that the government, a particular agency or a department might have procured multiple aspects of its back office or backend services through multiple service providers, but it is a de facto integrator. It's retaining the risk of if something goes wrong or if the way the contracts are set up, the uh, suppliers are not directly liable or accountable to each other for the handoffs. And so there is this challenge of finger pointing and very difficult, you know, to sort of point out and sort of distill exactly who is responsible for uh, the damage that has happened or the loss that has happened. So what happens, whether it was that particular Confederation line, of course they, you know, they, the penalties were paid for by the private concessionaire and all of that. But then the government also had to arrange for alternate transportation for people because this service was not working properly. So you know, so there are, they had to suffer revenue losses as well. The city had to. And so I think when we are talking about breakdowns in interfaces, you are seeing that, we are seeing that the government is absorbing a lot of those risks because the integration was not well captured in the, in the design stage. And so similarly I talk about exit costs and transition support, talk uh, about ownership and control of access to data and IP and knowledge. So if you don't foresee well in advance what you need to really own in terms of data IP knowledge. And it can swing m either ways. The government has a tendency to public procurement, you see this, that there is a tendency to want to over own like own more than what they need to own and at the same time under own so missing out on information that they must absolutely own. And what happens then is that your private concessionaire may just walk away with your IP when the contract ends. And you know, you don't have all of the historical data and knowledge that you, that helps you to make subsequent decisions uh, for your operations or your customers. Uh, so that's just walked away. And so again you have this problem of not being able to switch suppliers because easily because all of your data knowledge operating, you know, understanding all of that sits with the uh, vendors. There are additional, I think additional areas around. Now increasingly we see that technology is becoming a more critical, more and more critical component of physical infrastructure assets as well. Now technology moves at a much faster pace. So the asset life cycle is much shorter than any of the other probably most of the physical infrastructure that might have gone into your physical asset. So now you have to manage different asset cycles and you have to manage capital infusions around that. So again you have to make sure you have the right sets of obligations in place, right sets of provisions in place in the contract that identifies what is capital infusion or what is what is just a technology upgrade versus a significant capital infusion. So similarly around that uh, will impact the operating phase and down the road, down the stream has significant operation impact or commercial or even financial contributions right throughout the lifetime. Then I think there are some opportunities in contracts around. If you have really long term contracts, then your business over the course of the asset or uh, life cycle or over the course of the life of the public contract, you know, the costs might change your business as usual costs might change. Uh, some of the activities that were predominant, you know, they might become redundant over time. Things that were barely there in the beginning might in 10 years time become significant, significant cost. So you don't, you don't always know that and it's hard to predict that sometimes. So what happens is that generally you know, in the cost structure discussion that happens at the time of designing the deal, um, there'll be certain costs that are identified and everything else is sort of clubbed under some sort of change order mechanism. And so what one observes is that ah, over a period of time those unplanned costs start becoming bigger and your change order costs are really inflated, whereas everything else might have remained somewhat steady. And so you don't have a way of addressing it. You don't have a way of procuring the costs that you have in there competitively or marking them to the market because of the way that you've done the original deal. So these are the kind of risks that keep coming up and, and you keep seeing them across different uh, assets.
Evgenia: What's fascinating to me, I think uh, now I'm gonna try drawing in parallel. What you're describing is essentially a slightly different take on what we traditionally would have called the integration or interface management. But I think this is a bit of an, maybe not an evolved, this hasn't evolved enough to the needs that you're describing in terms of the future needs, the technology needs or AI needs of any asset. What I'd like to understand is your view on what should existing or develop projects and development. If we're being practical, how do we ensure that there is a responsible person party that would look at those things? Because I think it's, it's kind of one of those things that it's a gray area, there is less expertise. You often, what you're describing, you know, we've all seen where you have designers, you have contractors, you have O and M, M either private or public side, the end owner of the asset or operator. And they often do not fundamentally understand the needs of each other. They don't understand the long term impact. And it usually breaks somewhere in between, you know, final uh, stage stages of construction or early design. But then you're layering on additional elements of you know, customization and AI and digital technology. What are your thoughts on how do we make this a tangible, you know, maybe idea or solution to help projects identify how to deal with that? Like is that a role, is that a, you know, another contractual article? What, what is it? How do we name this problem?
RADHIKA: So I think some of this can be sorted contractually. I don't think that some of these decisions are um, you know, net new. It is just a different way of thinking around them. I'll give you an interesting example. Uh, there was a client that I worked for and they uh, they had an asset that um, they had, you know, a specialized physical asset that they had commissioned to meet their storage needs. And their contract was coming up for expiry. As for the terms of the contract, the day the contract ended, the client had to move to a different vendor. There was no provision in the contract that allowed for storing uh, the inventory for even a day longer. There was no provision for A grace period or some time period or a possible extension. And this was a fairly long term contract and public law does not allow you to go back to the same supplier that easily. So the client felt locked in because now they had to get a specialized service provider to build them a specialized asset by the end of the date of that particular contract. And also now because the vendor was not required to do provide any transition support. Okay. So now the client was also locked into the vendor. Vendor. If the vendor said it's going to take me, and the vendor originally said it's going to take us three years to move everything. Um, this is a huge inventory. Uh, when I went to the market, the market told me, well, it can be anywhere between three to six months. But now you're dependent on the vendor, you're dependent on the vendor support to move your inventory. And so the client felt, again, it felt that it had a very, you know, the client had a very compressed timeline to make that happen, to find another service provider to build them that capability within that time. Now exit strategy or you know, transition support. These are not, these are not net new. I'm not telling you something that's like really innovative or you know, novel idea. It's just where we are missing, where we are missing that long term thinking. So what it boils down to, I don't think you necessarily need a new body or uh, a different person who's looking at. Because it would be very rare that somebody would have lived a contract from its beginning to its end. You're not an asset. We're looking at 50 year assets, 20, 30 year assets, maybe longer. Uh, so it's a matter of industry practice and discipline. And so how do we build that quickly? So I think what we need is more, more multidisciplinary practitioners in the industry. We need leaders who can think across finance, technology, technical aspects of whatever that uh, asset you're building. They can think about risk from all of the different perspectives, the contract, the commercial architecture, um, all of those things. But in reality very often what happens is that when you're building these large infrastructure projects or long term sort of public service delivery contracts, the people who are managing the public entity that is managing these uh, assets, they're under a lot of pressure to deliver as soon as possible within certain timelines. And they are also tied to budget cycles, approval cycles, even political cycles, funding, political cycle, all of those things. Right. So you need leaders who can keep, who can get the job done in the short term, um, keeping the short term, um, pressures you Know, managing the short term pressures but also keeping their eye on the long term. And I think that's where some shift needs to happen as an industry that we have to start thinking about the long term consequences of what we are doing. Especially when it comes to, and we've been talking a lot about operational risks and financial risks, but especially when it comes to community side of things, when it comes to um, environmental side of things, the impact doesn't show up immediately. The impact always shows up over a period of time. So we need people who can think long term, who can think about the consequences of the decisions we are making from a risk, like a multidisciplinary risk perspective and impact, uh, perspective. But like I said, the real challenge is you've got a lot of short term pressures to get the thing, get the job done. So it's a matter of practice. Uh, we have to do exactly what we're doing here, which is to discuss and talk about our experiences, discuss and you know, and talk about where we feel some of the mismatches are or some of the gaps are, uh, so that you know, we can share information within, within the industry. Uh, uh, a third area where I think there is some gap I think is system level thinking. Right. Which is typically done at the government level.
Evgenia: I was going to say systems. Yeah. Mhm.
RADHIKA: Yeah. So are we doing like. We don't have a body that does integrated planning as of now. So we've got political offices or institutions who are developing different assets. Of course we have organizations that are looking at certain sectors, uh, for maybe housing or um, other areas. They're focusing more narrowly uh, on those areas. So there are bodies that are looking at specific sectors, but we don't have that integrated view of all of the infrastructure assets that we are developing at present. Whenever an infrastructure asset, especially a very large size one like a big toll road or a real, you know, a big subway, state subway or a big uh, national rail sort of uh, infrastructure airport. Whenever these kind of infrastructure assets are being developed, they will even data centers, that's actually a great example. They will immediately create cascading demands for other infrastructure. Right. Immediately there is demand for water, there is demand for uh, electricity or power, um, and other resources. That's on one hand you have cascading demand, on the other you have competing demands. So there has to be some system thinking around creating the right sort of, you don't necessarily have to sit and plan everything but creating the right conditions so that infrastructure investments can happen in a complementary manner. So you've Got, for example, you've got a major hub airport and you've got a city airport that could operate in a complementary manner as part of a regional aviation system. But then you have to develop the commercial conditions for that to happen, the policy conditions for that to happen as well, so that, that can happen. Otherwise they might compete with each other.
Evgenia: Absolutely. And what you're describing, um, to your earlier point about us sharing, um, you know, our experiences and kind of problems, and I think what it helps the industry as a, as a, as a whole is to really name the problem more clearly. Because I think articulating the problem and naming it is the first step to then finding out the solution. We're all kind of talking about similar things and I think your point. I'm so, as I've been doing this podcast for a while, I'm starting to hear, you know, themes and kind of the keywords that are coming up from all these different people that we interview that relate to the industry. And I truly hope we can one day make a greater impact by really identifying it and making something about it. But what you're talking about, um, just now about the economics and the policy, you know, having an integrated approach that also relates to procurement, that also relates to, to AI. What is it that traditionally we discuss these problems, but if we were to look back at all of these things that we, uh, these issues that we bring up, what needs to change in how public sector leads, this design of major programs or major projects to make it a long term, uh, investable decision or like, which comes first at the end of the day? Is there such, such idea that one has to come first? How would you answer that?
RADHIKA: Leaders have to come first. In my view. Leaders have to come first. I think we need more system thinkers, we need more multidisciplinary leaders, uh, who are empowered in the right way to make decisions also and who can hold complexity across finance, operations, technical aspects. So people need to move around the table a little bit more, you know, similar to what I've done, although that's not what I was going for, but was natural, see, But I see the benefit of that. Look at the banking sector, okay? Technology was always a support service. Banking, you know, customer engagement, all of this were the front end of the service. And technology was the back end of the entire sort of banking industry. With fintech coming in with all sorts of, you know, innovation that's happened in the banking sector. You have technology and retail bankers or even wealth. Now, um, I think people from across the different branches of banking, um, you see that people, the technology analysts and the bankers, they're working together to create products, they're working together to create services and improvise on them on a regular basis. So we've already seen that multidisciplinary move happen in many industries. I just cited one about the banking industry. But we're seeing that in insurance, we're seeing that in many other, certainly in financial services, but also some of the other industries. So why not infrastructure? What stops us? It is important that I think the future of work is multidisciplinary, it is collaborative. So we have to, I think it's very important to have leaders who have that sort of multidisciplinary experience. Experience and can bring it to the, you know, to their work. And uh, and, and also I think it is for all of us to sort of make the effort to continuously learn about the world around us, about whether it is, you know, successful, uh, or failed or whatever cases and keep the, that positivity around. Things can change and this is how we, we will change them, and this is how we will change them. We, we definitely need the, the right sort of attitude and values as well in the way that we work. The second thing, in terms of the more tangible sort of thing, we can change with AI. And um, I've discussed about this. In fact, there's a whole article on AI procurement, how public procurement of AI should be approached. And it emphasizes on having a separate discovery phase.
Evgenia: Yes, talk about that please. Mhm.
RADHIKA: So before we get into the uh, procurement phase itself, it would actually benefit us to have a separate discovery phase. AI is of different types, especially when it comes to agentic and uh, even generative AI for that matter, where you're giving the AI the capability to take decisions and execute on tasks based on the analysis that it's doing. There is a certain element of risk that you're handing over, right? So depending on the complexity of the AI capability that you're hiring, you need to understand where it sits in your operating model, how you can derive value out of it. And so it is very important that we have a discovery phase that helps us to understand for the government especially, what is the quality of data. We know government's sitting on a lot of data, A uh, lot of data. But what is the quality of data? Is it very fragmented? Is it sitting across different systems? Is it sitting across different suppliers? Can it really be used to make sensor? Can you actually derive clear patterns from that data? Or will it hallucinate? The other thing with, I think the standard Risks around government data, um, the sensitivity around it wanted to be exploited, uh, in any way. Those still apply. But also with government there are a lot of constraints that are security requirements, there are privacy requirements that are legal requirements and all of that. So when you're approaching AI, uh, vendor and saying, well, I've got these many constraints and I don't yet know what is the quality of my data, you know, you can't do the procurement in a traditional format, a fixed price sort of format, because you need to discover what can be done with the data you have within the constraints that you have. And so you have to give the vendor some access to the government data and the environment in which they're operating to be able to look through all of it and say, this is how workflows can be designed and this is how they can benefit you. That's one aspect as far as the design of the AI itself goes. The second aspect is around the operating model. We talked about it, uh, we touched on this earlier in the conversation. But you know, this, there is this, uh, AI is very attractive for everybody at the moment. It can do so much good, especially in public services. Um, the potential is very high. But how do you know whether adapting an AI or adopting an AI, uh, capability is actually going to deliver value and not cost? It's not an added cost. There's a lot of work that needs to be done, um, in understanding what is your operating model today. What is the true problem that you're trying to solve for. So if you're trying to solve, let's say, problem X, you think there is a, you know, you have this perception that there is a problem X that can be solved by adopting AI. But actually when you look into your operating model and you see where the operations are breaking down, maybe it could be, know, handover of, um, data, uh, or services between the different suppliers you've got and your own organization or your own business functions. Maybe it could be, uh, you know, the quality of the, in which, uh, you know, of the data itself. You might realize that AI is not the solution because you've not fixed everything else that needs to happen before, right? So that you can leverage the AI for its value. Uh, so that kind of exercise needs to happen as well. And then you need to work with the AI vendor to understand if you adopt firstly what the capability would look like. And then if you adopt the AI, how is it going to change your operating model. So, especially when we talk about generative AI or agentic AI, for instance, where you are independently you're basically telling the AI system I'm using very sort of simple casual language just to. But you're basically telling the AI system that I want you to analyze this, execute this, this, this, and then based on this information, I want you to make a decision and make the final decision and then maybe a human will review what you've done. Okay, Very simply put, now for the agentic AI to work, it may need to pull data from different systems, right? So that means your technology needs to be well integrated, your AI itself needs to be. So is your technology ready for that? Is your supplier system ready for that? What do you need to do in order to make that integration happen? Is your data standardized what you're asking your suppliers to give you? Right. Is there a huge element of human judgment that you would apply to the kind of work that you're doing? Is there a lot of sensitivity in the kind of work that you're doing? So especially when it comes to areas where the reason you're using the AI for is there is a huge risk, there's a lot of uncertainty around, um, the impact of what you're doing. There's lack of clarity around how it could, adopting the AI itself could transform your operating model, your people, processes like your governance systems, your policies and rights and all of that thing, even ownership rights. Because vendors, what AI vendors are looking to do is that they are looking to use the data and the work and the processing that's happening on top of that, take the learning away and then commercialize it by developing models on top of that experience. Right? So you don't want to suffocate the vendor from doing that. But then because you're government and you've got data sensitivities and all sorts of other sort of legal requirements, you might. So all of that can be sorted through in the discovery phase. And uh, it is important because that it's a paid discovery phase, because you're putting the time and effort and it can also allow you to say, well, you know, this was a great exercise, now we understand where we stand, what our problem is and what are we trying to solve, but this is not viable.
Evgenia: Mhm. Yeah. And what you're describing, I think most organizations would absolutely relate to and are struggling with just the sheer innovation side and the newness of everything that we're experiencing with AI and we're trying to apply it to all these different problems. And we're saying in some instances that technology and AI is the solution. But what you're describing is that we really need to look in the core of what the business operations are to truly understand how we can solve that problem. But sometimes there are limitations with data and access to certain processes that might be highly sensitive that um, block that innovation and development aspect of the technology that might be truly needed because they just don't have full understanding of that operations or the public side client.
RADHIKA: In this, in this example, yeah, I think that's a dry. Like I, um, and I, again, I've given some examples around this. So, so let's say there is a, ah, there is a government organization which manages, I don't know, 50,000, uh, public records of some sort. Okay. Or documents or confidential documents. And it is struggling with, or not struggling with, but basically at present most of its process is around looking for a document, finding the document, um, you know, putting it together, redacting confidential information from it, and then producing it to the customer. All of that process has been done manually or some of it has been done manually. So when you have to look through, search through 50,000 documents, that's a huge task for anybody to do it. When I say manually, I don't literally mean but still looking at it. But you know, there's maybe, yeah, maybe you have an index in an Excel form or something. Right. So it might seem like the most exciting solution AI can offer. There is that AI, you put, you digitize all of the documents of 50,000 documents. AI will search them for you, it'll develop the search terms for you, it'll scan all of them for you, and when you need something, you just put in a search term and whatever documents that are associated with those search terms will turn up and then AI will also redact them for you and AI will even send them directly to your customer. Such a fantastic solution. But then what if a lot of that data sits in the physical form and a lot of it sits in a digital form? So you have to do the digitization. There is a data problem. Or maybe everything might be digital, but the metadata around it or the descriptors of those documents, the index, there might be inconsistencies within that. So again, it's a data problem then, because you're dealing with public documents, official government documents, and I mentioned redaction. So some are very confidential. So who decides how do you release them? Who decides that? Yes, they can. You can create a digital record and uh, you can uh, disseminate certain types of documents to certain types of customers without, you know, human overlooking it or human judgment, etc. That's a governance problem, right? What happens to the people who are involved in doing that process now? What sort of training do they need to have in order to work through the new process that AI has now created for you? Now, do you let the AI send the document directly to the customer or do you go through it before you send it off? And for different types of documents, you could have different types of access rights. You may m change your revenue model. Previously, if you were charging 10, you know, $5 for a document, now you might say, okay, if I give you premium service, you know, I'm going to charge you 20, uh, like those kind of things might come up. So that's an operating model change. So you see, there is a. It changes the way you do your business because of what you're adopting. But again, if the underlying quality of the data, the governance issues, they are not sorted for there is no use, the data will not, the AI will not fetch you any value.
Evgenia: You truly have a systems worldview on these, uh, various siloed but yet integrated and interdependent issues that we're talking about. Fascinating. We're almost at time here. I can't believe it. Um, the time just flew by. Maybe as we wrap things up, I'll ask you, I have another kind of last question for you. What has your own career taught you about making good decisions? When the answer isn't obvious? We talk a lot about uncertainty today and kind of looking and predicting the future. What has own career taught you about making good decisions?
RADHIKA: I generally ask a lot of structural questions. I often start with the purpose. Why are we doing what we are doing? Who is it going to impact the solution? Like, do I have all of the potential solutions that could address the problem? And um, does the solution in my mind address all of the interests of the people who are not people, the parties who are involved. So I'm still sticking with infrastructure in my mind. So if I'm building something, can it hold commercial discipline? If I'm on the public sector side, can it hold fiscal discipline? What would it take for that to be true? And I think for me personally, I'm able to work through a lot of complexity. Uncertainty, uh, uncertainty, yes. And I think, uh, that's a byproduct of having moved around, like moved countries moved, you know, having to reinvent myself again and again. I think that is the one muscle you do build. So for me, I think it starts with what am I, what is my, what is my end goal? What do I really want to achieve and how do I do it? In a way that is fair and disciplined in a commercial way. Um, and, uh, of course, you know, it also has the right kind of impact. I generally think long term, backwards. Um, and I think there's some of the things that I'm talking about, you know, come from that perspective. Aligns very well with the infrastructure sector. So I think I've chosen the right industry for myself.
Evgenia: Absolutely. Well, this has been a very interesting conversation. Radhika, thank you so much for joining. I look forward to reading more of your articles and perhaps having you back next time to talk, talk more. Yeah.
RADHIKA: Thank you so much for, for reaching out. And thank you so much for giving me the opportunity to share my, um, my experience and my perspectives. I think this has been fun. So thank you so much.
Evgenia: Certainly. Thank you very much. That's it for this episode of Navigating Major Programs. If you enjoyed the conversation, please follow the show and consider leaving a review. To continue elevating the conversation, join us on the navigating major programs LinkedIn page. I'm Evgenia Jilina um, alongside Riccardo Cosentino and Shormila Chatterjee. Thank you for listening.