Quantum Computing Inc.

Hello, and thank you for joining our IBM Quantum Industry Webinar Series. Today, we are excited to share with you our webinar on how to become quantum-ready in finance. We are joined today by Vincent Beltrani, our quantum global lead in GSI and consulting partners at IBM, and then three of our service partners. We have Indranil Mitra, a VP at LTI Mindtree Research, Aaron Kemp, the US quantum lead for KPMG, and Evangelos Karamatskos, a global innovation quantum R&D leader at EY. So we're excited to hear what our service partners are doing in the finance space with quantum computing and how they're helping clients get quantum ready. So to just a few notes before we begin, you can feel free to ask questions in the Q&A dialogue box throughout the event. the question that comes to mind, feel free to go ahead and post it there. We'll be taking questions at the end. So feel free to add those there. This session is being recorded and we will share the recording with you after the event. And for those who weren't able to make it live, but who registered will also receive a recording later. So without further ado, I will pass it over to Vincent Beltrani and we'll get started today.

Great. Thank you very much, Jennifer. So thanks, everybody, for joining our webinar. My name is Vincent Beltrani, and I'm the Global Lead for Strategic Partnerships here at IBM Quantum. And my role is really to collaborate with our partners, our key service partners like UI, KPMG, and LTI Mindtree, who we'll all hear from today, to help them build out capabilities and scale quantum services to deliver to our end clients. I think just to set the stage a bit, our mission at IBM is to bring useful quantum computing to the world. And useful is really the key word, especially for an industry like finance. So we're here today to talk about how to get quantum ready in finance. And let me just give you a brief overview before I pass it over to the speakers. Just to set the stage briefly, quantum computing is not just about a faster version of what we already have today with our classical compute. It's a completely new way of doing computing that leverages really the fundamental laws of nature, quantum mechanics, to solve problems that might otherwise be classically intractable. Problems that are either too complex, too large, that they would even take the world's fastest supercomputers billions of years to solve. And that really opens up a set of entirely new possibilities across nearly every vertical. Just to look a little bit at the economics behind the driving force, we're seeing about $55 billion in global investment being poured into the ecosystem. And that's just not a future looking statement. It's really translating into real customer spending, growing at a 48, 50% compound annual growth rate year over year. These are really Moving from what used to be small experiments into more strategic significant bets for our clients. And what does that look like in practice? It really looks like a massive ramp up in applied research and in client work. I've seen a number of active proof of concepts grow. from on the order of tens back in 2018 to over 150, 200 today across all industries. So it is pretty clear that the race to build quantum capabilities and to secure a quantum advantage is clearly underway. IBM has been consistently identified as a leader in the quantum computing space based both on our current capabilities and on our long-term strategy. And what I really want to emphasize there is that a key part of our entire strategy at IBM Quantum is a transparent public roadmap. We have a strong record of continuously delivering against that roadmap. And that gives our clients and that gives our partners really the assurance that they can build their own strategic plans on our platform and start building on our platform for years to come. We're committed to that technology for the future of quantum computing. But we know that technology alone isn't enough. A new era of computing requires a completely new ecosystem. From the ground up, we've built the largest quantum community in the industry with over 600,000 registered users, over 300 organizations that are part of our IBM quantum network. And really at the heart of it, we're all driving this innovation together. So we have an ecosystem of industry leaders, of startups and research labs, which really is one of our greatest strengths. I mean, and crucially, this ecosystem includes the world's leading strategic consulting partners, right? Those are the organizations with the industry expertise who can really bridge that gap between maybe quantum potential and business strategy. And those are partners like EY, like KPMG and like LTI Mindtree that we're going to hear from in just a moment. So just a little bit of an overview. Together with our whole ecosystem, we're mapping quantum's impact across all major industries. As you can see from our slide, The problems we're targeting normally fall into a bunch of broad categories, chemistry and materials, optimization, quantum machine learning, potentially partial differential equations. The key here is that that applies to everything from developing new materials in aerospace, batteries for automotive and energy sector, and of course, to tackling complex risk and optimization problems in financial services. So just to dive in a little bit deeper into financial services before I pass it off to our partners, as you see here, those general problem classes, like the optimization class, the machine learning class, translate directly into some of the most valuable and complex challenges in the financial industry. We're talking about things like enhancing portfolio optimization, improving your risk prediction models, even developing whole new systems to detect complex financial fraud and anti money laundering. And that's a lot of the kind of what and why behind this. And what I want to do now is pass it off to our partners to start talking more about the how. How are our service delivery partners solving problems for our clients today? So with that, let me pass it off to Indranil Mitra, the VP of Research at LTI Mindtree.

Thank you, Vincent, and I appreciate this opportunity forum that IBM has stated to share our perspective on quantum applicability in the financial services domain. Just to give you a sense of how I intend to spend the next 10 to 15 minutes, I'll be briefly introducing the organization I represent, LTMITree, and share our journey in quantum as a service integration partner for our clients and prospects. And based on that journey, we formed a perspective on how quantum adoption roadmap would play out in the financial services domain. and share a little perspective on that and then we'll deep dive into some of the use cases that we've been working on. So LTM Entry is the leading IT service global system integration company based out of India. We are part of the multinational conglomerate Larsen & Tuchel Limited. In our own right, we are a $4.5 billion organization serving globally 700 plus customers operating in 40 plus countries through 84,000 plus associates. And the financial sector happens to be the largest industry segment that we serve. We have about 170 plus customers in the financial services segment. Many of whom happen to be the top global banks, US based, UK based and other parts of the world. So that has given us a lot of exposure to the business context of the financial services sector. And that has informed our perspective on how the financial services industry could benefit from the applicability of quantum technologies. Some of the areas where we've seen early successes, asset and wealth management, fraud detection, and data prediction, risk monitoring kind of scenarios. But I'll talk a little bit more about that in the next segment. A little perspective on our own quantum journey as a technology company. We obviously keep an eye on the emerging technology horizon and for the last. Over the last 4, 5 years, we've seen that quantum has started to make a transition from being. In the domain of research and. lab-based experimentation to actual industry experimentation and adoption. And that was a clear signal to us that as an IT service company, we should be starting to build a perspective and capability on quantum. So in 2022, we started our journey with quantum and now we have a team of 20 plus internal researchers and engineers who are actively exploring quantum. We have a body of work from a publication perspective But our real interest is in building industry-grade prototypes, which can be the baseline versions for pilot projects in the financial services and other sectors from a quantum technology applicability perspective. While we have been building our own capability in-house, we realized that it is important to be connected to the ecosystem because there is so much innovation going on across the ecosystem that it would be foolish not to be a party to that so that we can learn from our partners and learn where the success lies and which are the pitfalls to be avoided. With that in mind, we have tied up with a number of leading academic institutions in Europe, UK, India, and the US Americas. And we are also closely aligned as an industry partner to the government-funded research hubs in the UK and India, where we are a listed industry partner. Last but not the least, having the right technology partnerships is very important. IBM is a thought leader and vanguard in the quantum technology development sector. And we made the strategic choice to join the IBM quantum network. And this is the second year in which we've been part of the network. And that has added significant impetus to our own efforts in developing quantum capabilities and prototypes. As I said, while our focus is on research, it is very much industry aligned because that is where we see the real impact of quantum. We built a number of prototypes in partnership with IBM and a number of startups and academic partners in the area of fraud detection, health prediction, energy demand forecasting, etc. And as we speak, we are engaged along with IBM in the context of a leading PNC insurer to explore the applicability of quantum optimization in the way they handle claims in the ecosystem. So that's an interesting project we are working on. And another interesting metric I wanted to share with you before I move on from this segment is that in our annual customer satisfaction survey, one of the inquiries we made of our customers is that beyond the ongoing adoption of AI and cloud, what are the technologies that they see will have a significant impact in their ecosystem? And we were really happy to see that as many as 20% of our customers believe quantum will be significant in their enterprise ecosystem in the next three to five years, which is a significant uptick from where it was two years ago, which gives us added validation that quantum is starting to make an impact on mainstream technology practitioners' mindsets and is becoming very ripe for widespread adoption. So the typical journey which we have identified enterprises across sectors, including financial services, should go through would be to become aware on the possibilities of quantum, because as a technology, it is still evolving and, you know, and emerging. So there is a bit of hype. So it's important that you work with organizations such as those on this call or others outside to become aware of the quantum possibilities and then prepare yourself by conducting those little experiments as the technology matures so that you know which ideas would work for your enterprise well and which ideas may not be that appropriate for a quantum based implementation. so that when the time comes, you are not at a disadvantage, but rather at an advantage to adopt the scaled up quantum infrastructure to add value to it. With that background, I'll jump into the possibilities of quantum technology in the financial services domain. So we look at quantum computing possibilities across sectors in three broad vectors. first one is simulation now quantum simulation is especially promising in domains like quantum chemistry material science and drug discovery etc but there are scenarios in financial services industry as well which could do which could benefit from quantum simulation which are worthy of exploration like market analysis risk simulation asset pricing etc from a Quantum perspective, the next area of exploration that we are particularly excited about is the combined play of quantum and AI, wherein we found that with certain use cases in the machine learning ecosystem, there is significant advantage in using quantum algorithms when it comes to machine learning use cases when the data set is of particular nature. And we found some good success, empirical evidence of good results in quantum ML vis-a-vis classical ML. in the financial services domain and I'll talk a little bit more about that in the subsequent slide. Quantum optimization obviously is a strong area for exploring possibilities across domains. We've done a lot of experimentation with portfolio optimization and I'm happy to see that some of the early experiments have been positive results in terms of portfolio optimization. And there are a number of other scenarios as well, like balance sheet optimization, etc., which are worthy of exploration in the financial services sector. Last but not the least, I would be amiss not to make a reference to the threat of quantum, wherein, as many of you would be aware, quantum computing scaled up to a high level of scalability, could power algorithms that could compromise existing cryptographic algorithms like RSA, etc., So it's important that enterprises in the financial services domain, which is extremely data sensitive, should start exploring quantum safe solutions very actively. I would now like to spend time on a couple of scenarios of quantum use cases that we've worked upon. The first one refers to credit card transaction fraud detection, which is a use case we had an opportunity to work on. We received some anonymized data from a credit card company provider out of the Asia Pacific. And we found that the data had a specific quality, which was very suited for quantum machine learning algorithms, which is that the data was highly imbalanced in the sense that fraud occurrences were low compared to genuine occurrences. And the data was very feature rich, so it was highly dimensional. Given this situation, we found that Quantum support vector machine algorithm was able to perform better than classical support vector machine algorithm, as you can see in the slide. The prediction accuracy was higher and false positives and false negatives, so a significant reduction, which gives us reason to believe that quantum ML is an area which should be explored actively by the financial services sector in the near future. And we are working with IBM on building a set of tools which will help classical data scientists and machine learning engineers explore the quantum algorithms without having to worry about complex concepts like data encoding or dealing with the underlying hardware. So our effort is to make quantum machine learning more accessible and democratized across the ecosystem. Another use case I wanted to talk about was quantum safe assessment. As I mentioned, with the onset of large scale quantum computers, powering some of the algorithms like Shor's algorithm, existing cryptographic protocols may be vulnerable to attacks. And in that context, we are advising more and more of our financial services customers to start conducting assessments to take stock of their cryptographic bill of material so that they are aware of what their cryptographic footprint looks like and be prepared to remediate the same with quantum safe protocols. We have partnered with IBM using a couple of their very advanced tools like IBM Quantum Safe Explorer and IBM Guardian Quantum Safe Explorer which help us assess the cryptographic bill of material at an infrastructure layer, at a network layer, and at an application layer, and help customers come up with a strategy to remediate the most susceptible areas right now. Because financial services is a sector which has a very sensitive nature of data, and the data has a long shelf life, which means that it is susceptible even today to harvest now decrepitated kind of attack schemes. And that is why it is important that financial services organizations start assessing their landscape and start building their remediation and agility strategies to be more crypto safe. With that, I would like to conclude my presentation because the next segment is going to cover additional details and hand it back to Jennifer for the next segment. Thank you.

Thanks. Yes, we can go on to Aaron Kemp. Aaron, feel free to take it over.

Thank you very much. Thank you everybody for coming today. Thank you IBM for having KPMG out. Our partnership with IBM is in its third year now in the quantum space. And we've really begun to see traction with both IBM in our research and our clients. Looking forward to discussing a bit of what KPMG has going on in the space, what we're seeing with our clients, what we're seeing with industry, and why we think quantum is going to be a key part of the coming years. I'm the U.S. Quantum Lead for KPMG. I've spent about the last four years building out our offerings to support our clients in this burgeoning space. Quantum is rapidly becoming the topic du jour. It's moved from research into the main news cycle. Organizations are really beginning to question where they need to be playing in it, what response they need to have coming to the guidance that's coming out, the regulations, and where their strategies should align. We'll go through, start with the opportunities. There is a good and a bad side to Quantum. We're seeing a lot of research going on. There's been announcements in the last few months about the financial services and opportunities that are actually coming out in the space utilizing Quantum. We're involved in some of that research with our clients. But the first thing to note is Quantum is growing. We're seeing significant growth and significant projections of growth over the coming years. That 2040 number, I've actually seen new reports now that it moved that to 2035, and it is now $1 trillion by 2035. Organizational investments are increasing. Organizations are starting to realize that beyond the AI industry, Being the AI hype is a new form of computing coming down the road that will offer opportunities, and they're beginning to invest. They're beginning to look at what they're going to need staffing-wise. They're beginning to project where this is going to play strategically within their organization, and they're starting to set aside that funding to do that. This is also occurring at the nation-state level. We've seen significant investments in North America. We have a really good startup company We're seeing lots of different types of research on different types of quantum systems, and we're really beginning to see the monetization of those research projects. It's moving from the lab to things where investment's going. Anybody keeping track of the quantum stock investments has seen the significant changes, people looking to really grow companies, and we're seeing some very significant changes valuations of startups. runs coming up for some of these startups. There's been some large investments regionally between things like Chicago Quantum Exchange, the Elevate Quantum project out in Colorado and Wyoming, and we're really seeing a consolidation of regional areas trying to build quantum hubs. China has continued to invest significantly into quantum, continues to grow. We've seen about $35 billion in investment in the last 10 years. There'll be launching some quantum key distribution satellites this year. They're very much researching down the quantum networking space, and we're really seeing a drive to take the lead in that space. So right now it's kind of neck and neck between us and China. Europe is also investing largely. Australia also has a significant quantum program. So we're seeing nation states really play in the space as well as startups and industry. Use cases we're seeing across this, we are actively working with IBM, our partners for research for clients. These are some of the places we've begun to actually take a look at where quantum can provide an opportunity for our clients. Fraud detection is one we're actively working right now. We're researching the ability of quantum computers to pick up smaller indicators in fraud from data sets for credit cards, hoping to be able to reduce the number of false positives and really begin to pick up those little indicators that show fraud is occurring so the applications are more accurate and we can push that cost down. Portfolio optimization continues to be a big area of research. There's been some announcements around that recently with some significant numbers. That continues to grow. We're pursuing that. The risk analysis Monte Carlo simulations are kind of the bread and butter of quantum systems. They're keyed up well to be able to do those types of problems, and we're pursuing that on numerous different routes. Then derivatives pricing. We think that we're going to be able to do more accurate pricing in the derivative space and really begin to take advantage of Quantum's ability to handle multiple variables and bring back more data that is useful in a more timely manner. The industry sees a lot of opportunity in Quantum. KPMG sees a lot of opportunity in Quantum. We're pursuing that to ensure our clients are receiving the support they need and that they are ready for this kind of new era of quantum computing. Unfortunately, with quantum more result of quantum than what result of quantum than quantum itself, there is a downside, which is the post quantum cryptographic migration. Shor's algorithm, Grover's algorithm, the ability to crack RSA, ECC, Diffie-Hellman, and the other algorithms that we use for our public key encryption is becoming a significant issue. NIST has produced the first three standards. Those came out in August of 2024. Those standards are set to replace our current cryptographic standards. So the organizations and financial services have really began to look at what that risk is going to be and how it's going to affect them. And we are positioning ourselves, KPMG, to support that migration. We have a product and a framework called QPREP. It's designed to walk an organization through this migration process. providing a roadmap and really spending the next three to five years getting the organization ready for post quantum becoming quantum resistant and being able to be cryptographically agile in a new space just to give you an example of what financial services are facing in the coming years These questions that you're seeing here on the gray background are from the FDIC-OCC bank exam this year. And these questions went to the board, really asking about the organization's posture and strategy towards quantum, what they were going to be doing, how the board was being educated. Were they vulnerable to quantum attack? What was going to be their responses? What are their plans to get through the next few years as we begin this migration process? So a lot of work being put into a lot of work and questions around that post-quantum cryptographic posture coming up. Organizations are beginning to look at what it takes to migrate. The EU has come out and recommended that all nation states begin their cryptographic baselining next year. That's the... That's the going into your systems, determining where the algorithm's at, what the algorithm is set to, what it supports, what its dependencies are, and really building that cryptographic bill of materials so that the organization understands their cryptographic posture. This is very different from how organizations have done cryptography for a long time. We tend to make the key bigger, push cryptography into the background, obfuscate it, and then we don't talk about it again. So this is a cultural change as well as a significant technical change. challenge organizations are going to have to not only migrate their cryptographic processes but they're going to have to work their vendor dependencies they're going to have to understand their technical debt replacing equipment that can't be upgraded to the new algorithms understanding where they're going to have to spend money um as well as become more cryptographically agile and for those that aren't familiar with cryptographic agility there's a belief that NIST believes that in this new post-quantum world, we're going to have to have a library of different algorithms based on different mathematical formulas in case something does happen. We don't have quantum computing at scale. We don't know that all the algorithms that have become standards will hold up. They're believed to be quantum resistant, but there is a plan within NIST to ensure that we have different mathematical formula to protect ourselves in case one of those does get broken. And several of them were broken as we were going through the testing process. So PQCs become very important. Financial organizations have a lot of data, a lot of transactional data, a lot of information about their users that has to be protected for significant periods of time. We've heard about harvest now, decrypt later. So this is all occurring now. Guidance is coming. The FS group, FSISAC, has produced a significant amount of data and information about what this transition is going to look like. The post-quantum financial infrastructure framework was just released, and that is really starting to drive the financial services towards this migration phase. They're starting to really break up what they need to do, getting into the early parts of their identification, their inventories, and looking at what the next three to five years will be. We're recommending through our Q prep that first we scope out what needs to be done, and then that initial asset inventory, creating the C-bomb. Moving into risk and data governance, this truly is a risk and data governance problem where we're trying to protect the right data with the right level of encryption and ensure that that is protected for the right amount of time. There are three standards right now for solutions exploration. There will be a fourth and fifth standard coming from this. So understanding the solutions that are available, how they work, how they scale is going to be a key part of this. Then finalizing the transition plan development, getting an implementation and remediation plan. This is going to be a difficult process. And then obviously down the road, getting into continuous monitoring to maintain that cryptographic agility and really be able to push through this new quantum era we're in and make sure that our transactions stay secure. So lots happening in the space. KPMG is definitely there to help. We are very happy to be a partner with IBM, both on the optimization side and on the remediation side with PQC using the Guardian suite. And it's just something that all of us are going to have to do as we move into the post-quantum world. So I am going to pass off to the next. So thank you.

Thank you very much, Aaron.

Oh, yeah, I just was going to pass to you, Evangelos. Thank you so much for joining. And yeah, we'll transition to you.

Thank you very much, Jennifer. So hello, everyone. I'm Evangelos Karamatskos. I'm the quantum R&D leader for EY Global Innovation. And I'm very excited to be here today and to share an overview of what quantum technologies at EY looks like. And first of all, I would like to thank IBM Quantum for organizing this webinar and for inviting me. And in particular, many thanks to Vincent Viltrani and Jennifer Janicek. Here's a brief outline of my presentation. So I will give a brief overview at the high level of what quantum technologies at EY looks like. And then delve a bit deeper into our quantum safe program, quantum computing use cases for finance. And finally, very briefly touch upon quantum strategy and risk where we have a framework developed at EY. Finally, I will have a quick summary where I really want to give you the key takeaways of my talk. So let's jump straight in. So EY is one of the largest professional services firms globally and operating globally in over 150 countries. And we are offering a wide range of services across various industries. Um, and traditionally we have also very strong presence, uh, and, uh, as trusted partners and financial services. So, um, our quantum technology initiative focuses on accelerating quantum adoption and helping our clients solve real problems. Um, and generate value through quantum technologies. So we are partnering in for doing this. We are partnering, for example, with industry leaders like IBM quantum and others. we are developing custom solutions that are tailored to our clients' needs. And that ranges from use case development and quantum algorithmic development to things like strategy and implementation roadmaps and other things. So we have expert teams across the globe addressing all pillars of quantum technologies, as you can see here. So this includes things like quantum computing and simulation, quantum communications, which has... in one sense, classical part, which is the post-quantum cryptography that we already heard about, and I will also talk a bit more in detail about, but also the quantum-resistant network side, and, of course, sensing and metrology. But, of course, there's also something more, which is a general enablement, which is things like setting up teams, raising awareness, training, upskilling people, the workforce, and things like that that we are doing. So on the left bottom, you can see some selected use cases. We have worked a lot with partners like the NCSR Democritus Academic Research Institution and IBM Quantum on developing quantum algorithms in the domains of quantum AI and optimization. And we also provide advisory services in quantum strategy and implementation roadmaps. But we also place a very strong emphasis on quantum safety together with our cybersecurity teams. So in contrast to Aaron, who started with opportunities, I would like to start with the risk side of things. So we have already heard about this, Quantum offers incredible opportunities, but on the other hand, there's also serious risks that one needs to consider. And traditionally, one of EY's core areas of expertise is related to risk, be it, for example, financial services risk management, tech risk, or the assurance-related risks. So I will start today with the quantum risk, which is the risk to the security of our digital infrastructure. So we have already heard about this, but quantum decryption might be only five years away, maybe 10 years. We don't know exactly the timeline. But the important point here is that this is already a pressing issue today and not just a future concern. And there are several reasons for that. One is the famous harvest now, decrypt later. And you can see on the upper right side on the graph that there are a lot of assets, digital assets like data, infrastructure, and other things that have a life cycle that is much larger than the anticipated time until quantum decryption becomes possible. And that's why governments, regulators and security experts are proactively addressing this threat by, on the one hand, raising awareness. So you can find a lot of thought leadership material, reports, potential migration roadmaps and things like that, that are out there published by very serious people. On the other hand, we have seen also development of quantum secure algorithms and standards where NIST, the National Institute for Standards and Technology, has emerged as a guiding beacon and last year published their standards. Of course, this is still work in progress, so there will be more. And we have also seen a lot of regulators interested really in this topic and discussing about how to move what needs to be done to ensure a secure transition to post-quantum cryptography. So the key takeaway here really is that organizations must act now for several reasons. The one is the one that I already told you about, the harvest, not the crib later, and the life cycle of important data and assets. Additionally, of course, transitioning to an agile post-quantum cryptographic environment requires a lot of time, and it also presents complex challenges. So this is not something easy. It will take time, and it's better to start now and have the time to do this properly. Another thing is that lack of PQC compliance will pose, this is what we predict will pose in the future, significant reputational and operational risks. So if everyone transitions, then one risks lagging behind. And, of course, cryptography is something that connects people. So this is for communicating with others. So if everyone transitions to PQC, it's difficult to be able to communicate with those people if you are not doing it too. And lastly, early adoption. So at EY, we've established a quantum safe program and post quantum cryptography lab to help our clients navigate this transition. Key steps here are in order to minimize adoption friction are to include to do crypto inventory, enhancing cyber agility, creating strategic roadmaps. This is very important. You need to know where you want to go and how to go there. So this is really one of the key components you need to do. And also to identify really the critical data and assets, because you can't do this transition for everything in your organization at the same time. So you need to prioritize. And for this, you need to understand exactly to do the inventory, understand your critical data and assets, prioritize and see how you get started with this transition. And really, one of the most important points here is crypto agility, which refers to the ability to be able in the future to withstand new risks that appear and really quickly mitigate those risks. And in order to do so, one needs to implement a crypto agile cybersecurity framework, which is something far from trivial. And one reason for this is that in the past with RSA and other cryptographic standards, people were so sure that no one will be able ever to break this encryption that people really did not implement this in an efficient and modular way. So we are assisting our clients in conducting inventory assessments to identify systems at risk and the same for data and other assets to evaluate the associated risks and prioritize the assets for migration. And finally, of course, we also facilitate the actual migration to post quantum cryptography standards, offering also afterwards, of course, monitoring and other services. So now moving away from the risk side, let's look at a bit more into the opportunities. So I have here a list of some use cases. You have seen some of them already earlier. So those are really use cases where we believe that financial services will benefit massively from quantum computing already now or in the near future. And what one needs to understand here is that quantum computers will enhance and not replace classical computing. But quantum computers will be really an add-on able to tackling computational problems that are extremely challenging and often impossible or out of reach for classical systems. So it's also important to understand that quantum computers won't be able to solve every computational task more efficiently than classical computers. There are specific categories of computational problems, which are believed for quantum computers to be well suited for quantum computers. And these include things like quantum AI, like pattern detection. We have already heard also about this. Optimization challenges like combinatorial optimization and problems in stochastic sampling and really generating true random numbers. In finance, it turns out that the mathematical structure of many computational problems aligns well with the capabilities of quantum computing, making exactly these use cases here to benefit from this technology. And we have very famous cases like fraud detection and anti-money laundering, which will benefit from quantum AI. I will discuss in detail a bit more in detail the portfolio optimization where we have worked a lot on that EY. But there are also other things like Monte Carlo simulations, for example, where you can really show that By using quantum random number generation and sampling from quantum distributions, you can get faster convergence and also better get, let's say, more complex distributions that have fat tails. And for example, the black swan events, you can have a better sampling of those events than classically. So now... In general, at EY, quantum optimization is one key focus area within the quantum computing initiatives. So we have created an optimization framework that facilitates fast prototyping and implementation of new optimization use cases. And this also provides access to various classical and quantum solvers for hybrid quantum classical optimization and enables execution across different quantum computing modalities. So this was the aim of having this platform where you can really access different machines and quickly prototype some problems and do some benchmarking and see how you can best solve this. Portfolio optimization is one such case and in particular portfolio optimization is the process of selecting the best distribution of financial assets while maximizing on the one hand the expected return and keeping some specific risk tolerance or minimizing at the same time the financial risk. And in general, this is an NP-hard combinatorial optimization problem, which is very difficult to solve classically. So if you have a basket portfolio of assets and you add one more, this scales very unfavorably. And on top of that, typically, people when doing portfolio optimization look really at the mean and variance models of the classical Markovits portfolio optimization. But if you look at real distributions, they are not perfectly Gaussian. So you have typical distributions that have some skewness, some asymmetry. And by using this mean variance optimization, you cannot really capture this asymmetry. What we have done here is that we have run the classical portfolio optimization and on top we have extended this to the MVS, so mean variance and skewness regime, where we also incorporate third-order moments of distributions, and this captures much richer risk profiles. And in particular, I want to show you some very recent results. So we have done this with a partner of us, QCI, and run this on their Dirac-3 analog quantum machine. And it's very interesting to see that On the one hand, we get very good results. So if you look at the Sharpe ratio or the max drawdown, so the Sharpe ratio is a measure of the return given your risk. So the higher the Sharpe ratio, the better. And the max drawdown, you typically want to have it as low as possible because this tells you, like, if your portfolio increases in value, how much is the maximum drawdown that you have from this point? So it's a measure of volatility. And we could show in this example that we have excellent runtimes, much faster than the classical case. We get very good results. But in addition to this, there's also one other point that's very important. And the point here is that if you use classical solvers by adding a third order moment, you are entering a non-convex optimization problem. And typically what you do classically is that either you develop some approximations and heuristics for this, or you linearize your problem, or you run some solvers that capture only part of the known convexity. The nice thing about the quantum solution is that you can natively incorporate these higher order moments into your optimization problem and run those without the need for every different problem, different portfolio, different constraints. to develop a separate optimization solution. So this is one of the big advantages. So finally, just a few words about our quantum strategy. So we developed a framework as a service offering to our clients to support them in their quantum journey and have them in developing and executing their quantum strategy and implementation roadmaps. So We think it's very important for organizations, particularly with emerging technologies like quantum, where it's often not obvious and you can't really predict where you are moving, what are the benefits, what you should do next, to have really a clear roadmap and to be able to measure, while you are executing your roadmap, the impact on the organization and the benefits. So the key offerings here are really to have a holistic view on organizational adoption across the organization, a clear quantum strategy and implementation roadmap that is adaptable to the needs of the key goals of the organization, and to understand where the key focus areas are, where one needs to focus in order to achieve the goals and manage the risks effectively. So with this, I want to summarize and conclude and really give the key takeaways for today. So the key takeaways are that quantum technology is really advancing at an unprecedented speed. It's really a very exciting time to be in quantum right now. In principle, every week we see announcement of big and significant achievements. And quantum advantage is really just around the corner. So we believe that maybe within the next year we will see some proven quantum advantage for some specific use case. There exists already today some claims of quantum advantage and they might be true, but of course they are still subject to scrutiny from the quantum ecosystem. So in my opinion, it's a bit too early to tell really that we have achieved this goal, but I think we are very, very close and maybe we already have. The ecosystem in total is growing with substantial public and private investments. So this year already, the investments have been in a single month and the quantum investments have been larger than the complete year before. And this, of course, helps the technology to advance and to mature much faster. Quantum adoption, whether addressing the risks or seizing opportunities, requires time and early adopters will gain a competitive advantage. And one of the reasons is that really solving actual problems with quantum computers requires deep expertise and experience, And you need to really be able to create tailored solutions to your problem and solve this problem. And it's very different from other technologies where you can just download some library or do an API call and call some solver and have a very black box kind of thing. Maybe in the future we will be there. But at the moment, really, you need people that are really experienced in developing quantum solutions. Then I talked about quantum cybersecurity. This is really a topic that is at the focus of what EY is doing. We want really to raise awareness across industry and institutions that immediate risk mitigation is required here. And lastly, quantum technologies encompass a wide range of innovations and technologies. It's a wide technology stack, and those are really driving progress across industries and the entire value chain. And it is crucial for us and for everyone who wants to really benefit from those technologies to look beyond the hype to really understand what are the real risks, what are the real opportunities, and how do I need to move and what will be my quantum journey to benefit from these advancements. So with this, I want to thank everyone joining today's call. I want just to remind you that you can download the two white papers that I've shown here in my talk from the materials on the platform. And please feel free to reach out to us if you would like to discuss quantum. Thank you.

Thank you so much, Evangelos, and to all of our speakers. I'm going to skip ahead a few slides here to get to... Yes, a few places you might visit in addition to the resources that are provided on your webinar dashboard, which you can definitely check out the downloadable resources provided by our speakers today. We also have IBM Quantum Business Foundations, which is an open course where you get a Credly badge. after taking it, where you can explore quantum readiness, and, of course, go to IBM Quantum Platform, where you can get started programming on real 100-plus qubit quantum computers today. But let's open the floor for questions. So I welcome all of our speakers to, you know, rejoin the discussion here. We have lots of good questions. Let's start. There was a question that had to do with, you know, We're seeing some instances along this path toward scientific and empirical demonstrations that might be toward quantum advantage, where people are starting to look at actually using production data. How difficult is it to integrate quantum into existing systems. Maybe you could talk to us a little bit about the steps that organizations can take today to start to think about that organizational readiness, the capabilities they might need to build to be able to do that, and just the steps required for getting positioned for production demonstrations.

Yeah, I can take that one, Jennifer. It's a really good question. So I can give you my perspective on the way I'm seeing a lot of our clients starting to see these early value signals and saying, how am I going to be able to start implementing this in my production systems, for example? So I can give you, here's an example of something. A company works on a machine learning model for evaluating credit risk, for example. They build that model and then they say, OK, let's move this model into production, which might mean in the early days running that report every night, evaluating all of your classical risk models and then making a call out to, you know, maybe through Qiskit or something, making a call to go and evaluate that from one of our quantum models at the same time. It seems like kind of an early step to have a model running in production that gets called every night. But I think what's important about things like that is that it introduces quantum ideas to your DevOps team, right? They need to understand how to go and deploy these types of things. It gives the team that maybe manages your CICD pipelines a chance to get comfortable deploying things into production. And it also helps your SRE teams, your site reliability engineers, and those teams evaluate what it's going to look like. So from today, I see kind of small migrations of individual POCs and pilots slowly making their way into production and then having our clients kind of get comfortable and get accustomed to what it's going to be like to run these kinds of codes in production.

Thanks, Vincent. Other thoughts from our speakers about what challenges organizations are facing when they're starting to think about this or opportunities and steps you all are helping them take along this path?

I would echo Vincent's points because ultimately quantum will coexist with classical paradigms of computing. So it's important that enterprise architects and thought leaders start imagining how their ecosystem would comprise of a hybrid sort of setup, wherein quantum and classical workloads feed off each other and provide the transformative outcome. So yeah, I would agree with Vincent that we need to think about architectural constructs wherein a classical workload hands off to a quantum workload and then takes it back and takes it forward to conclusions. So that's how we have to think about solutions going forward.

Yeah. And if I may add here, I totally agree with all what has been said. And in general, I mean, quantum is not something that will exist by itself. It needs classical compute in order to for things like quantum error correction, data pre-processing, readout, storage, things like that. But of course, we have also seen things like the quantum-centric supercomputing facilities coming up, which means that you have, in principle, a new paradigm of distributed computing, where you add to your GPUs, CPUs, TPUs, whatever you might have in your supercomputing facility, You had also your quantum computer where you can really offload specific parts of problems that can't be solved classically, efficiently. And in that sense, it's important for everyone who wants to be able to use these technologies beyond the skills and expertise that is needed, that there's an overall digital maturity that you need. So we are talking here about data management systems. We're talking about cloud access. We are talking about data governance and everything that comes together for mature digital infrastructure.

Clients also have to be aware that everything is not going to be quantum. As they get into the quantum space and they start to explore that part of the journey and learning where quantum will be applicable and where it won't be is going to be just as key as to, hey, this is a quantum problem. The experimentation may show that. the classical methodology used, high power compute clusters, et cetera, provide a better opportunity than quantum. And that is going to be an important thing for the organization to develop into of our experiments have proved we don't need to go down the quantum path on this particular topic because anybody in the space knows it's very niche where quantum is advantageous. And in a lot of times, the high power compute, once they re-optimize, catches up. So it's going to be a unique space where just because you have a quantum team, they may actually prove that the classical methodology is better.

Yeah, great point. Okay, to sort of a follow up question and addressing one of the other questions asked in the Q&A, is there a sense for how we know that certain types of data might be beneficial for exploration with quantum computing? Especially with machine learning, where a lot of it is really data set dependent.

Yeah, Jennifer, I can take a first stab at it and invite others to add on. But basically, our empirical experiments show that when the data is extremely unbalanced in the sense that the outcome that you're trying to predict has fewer occurrences and it's highly dimensional. Quantum algorithms have a better performance when it comes to training themselves on such data sets. And this is empirical data, empirical evidence so far, but I think a lot of exploration has to go as to what data complexities are best suited for quantum encoding and quantum training so that the inference quality improves. But empirical evidence indicates that when data is highly imbalanced, and region features, quantum support vector machines have a better training track report.

Right. So, Ranil, are you done?

Yeah, I think he muted.

Okay. Sorry, I didn't want to interrupt you. Yes, I mean, there are indications that for specific problems and machine learning algorithms, there is or there will be an advantage. And that might be that, for example, being able to detect patterns with the same accuracy with less data, training data, or imbalanced data, as Indranil mentioned. Or, for example, also, that's also one interesting thing that this... been researched in how far, for example, correlated data can be used in such algorithms. Another very interesting point, I think, is the generation of data that can be used for training AI models. So this is also something that some people are pursuing because, in principle, there are some mappings in the Hilbert space in quantum mechanics that you can't have classically. So you can generate data that you classically can't. So you can generate synthetic data that you can use for training AI models. But besides only the AI aspect, of course, there's also the aspect of sampling from distributions. And I think there has been recently in the past year proved by some big financial institutions, actually, that when you use quantum random number generators and you sample from distributions to quantum monte carlo you can really speed up convergence and you can better sample parts of distributions that are far out of the center that classically are not so accessible and i think this is also a very interesting point that you can really better sample those data and model those data

Great. Well, I know we only have a couple more minutes here, so we'll have to follow up with many of these questions offline, which we're able to do. I think maybe one final wrap-up question would be a good one to speak to. You all are members of IBM Quantum Network. Maybe you could speak to your, you know, the benefits that you've experienced of being in the network and, you know, the collaborations and how it's kind of furthered the work that you're able to do with clients, just so everybody maybe knows more about what the ecosystem is like.

Yeah, I think for KPMG, we're coming up on our Well, coming up on the end of our third year here, we've gone from being very nascent in the quantum space to running quantum research projects. I think the support from IBM, the engineering support has allowed us to accelerate our program internally. And honestly, the relationships between IBM and the teams we work with has gone on to support clients. So for us, it's been advantageous in growth internally and then growth externally. So it's been... It's been a great partnership and being able to get the quantum knowledge early as we go to the partner forums and we get to understand what's happening first so that we can get to our clients and have kind of that ecosystem lead has been key in moving us forward.

Yeah, I would echo what Aaron said in the sense that access to the ecosystem has been extremely beneficial to understand what's working and what's not working across the community. And that's been an extremely high value at point. Access to expertise from IBM has been a huge benefit. It has helped us fine tune a lot of the research we conduct in our organization and has improved the quality of our pilots by a huge level. And last but not the least, being a part of the network gives us access to the latest fleet of hardware, which helps us experiment that much more effectively. And the joint go-to-market support is extremely beneficial as well. So a multi-pronged partnership has been extremely beneficial for us as a member of the quantum

Yeah, so similarly for us, I mean, EY and IBM have a longstanding alliance partnership, and with IBM Quantum, we are also now at the end of our third year, and it has been nothing but exciting. So we have collaborated on very exciting projects together, and we got really great support and insights and inputs that helped us really understand better how to use the hardware, how to solve actual problems, and really improve our algorithms that run on IBM quantum hardware. And of course, like the people we have worked together, that's really great, great collaborations, just fun and getting nice work done. So thanks a lot for that.

Thanks, everyone. Yeah, wonderful to hear about the experiences and the collaborations occurring across the ecosystem as we all work together toward demonstrations of quantum advantage and building toward fault tolerance. So great to hear about the use cases and work that you all are doing. Lots of good questions in the Q&A box about you know, the heuristics with QML applications for finance and how organizations should be thinking about that versus, you know, areas where there's more, you know, formal proof. So we'll follow up with a lot of those offline. And of course, the replay will be made available to you all as well. So you can go back and And look, if you didn't catch anything and the slides are available for download as well. So thank you all for joining today. We're thrilled to have you with us and we'll look forward to seeing you at the next session.