10991_AI_Business_School_Ganesh_Padmanabhan_FINAL

[AI Business School] [What is changing in the financial services industry] [and why is AI important to consider in this sector?] >> Ganesh: The financial services industry is going through a massive transformation. The traditional bank that did everything from investments, loans, marketures, are now being disaggregated into multiple specific ecosystems. A few things to note: There is a new generation of customers and prospects that have fundamentally different expectations than prior generations. If you look at veld management, there is a $30 trillion veld transfer that's going on from baby boomers to younger generations. And every time this has happened in the past, clients change advisors. Less than 50% of the mass affluent market has life insurance. This new set of customers and prospects have fundamentally different expectations of engagement, and using AI to understand what these customers want, need at a very detailed level, will fundamentally transform the way you interact with them. Second, there is more data available than ever before to make investment management decisions, and it creates an opportunity for institutions to create new products, new asset classes and new digital experiences. There's a company in London called Bridgeweave, who is building this next generation of financial assets. They process 350 million data points, 135,000 calculations a day, to provide very specific insights on traded assets to individuals, asset managers and advisors. Technologies like machine learning and natural language processing can massively improve the productivity of employees within institutions. Banks and exchanges are using AI to do trade and market surveillance to reduce false positives and improve compliance, making the life of the trade compliance officers much more effective, much more productive. The opportunity with AI is an increasing revenue growth, reducing systemic risk and improving productivity, and organizations that take advantage of this will then in this new world in this new future. [What have you observed about how AI changes] [the way employees work in the financial sector?] >> Just like any other disruptive technology that changes and transforms markets and industries, even AI will have an impact on the labor market in general. But what we have seen happen is most of the -- even though the narrative for AI has been primarily written by Hollywood. When you think of AI you think of the Terminator, the big red button, the Ex Machina, the whole fully-automated world that is completely taking the humans off the loop. The more practical part of AI that is disrupting and changing industries is where you augment the human in the picture. So that leads to two different distinct patterns that we see. One, your employees in your organization, their daily life gets a little bit better. When you're a veld advisor who is actually supporting and handling 20 to 30 clients today, with AI, you have the opportunity to improve your productivity so you can now handle about 100 to 150 veld clients. The average day becomes a lot more fulfilling and it becomes a lot more productive for the employee with AI. If you look at 15 years ago, whether there was a role called social media marketing, it didn't exist because the internet created new industries, new patterns, and it created this new set of jobs that came to life. Similarly, with AI, you're going to start the creation of what I call a new college ops, where you're bringing out the best in what humans can do. And it's not about humans versus machines, but about machines and humans. How do they actually work together to accentuate, to improve, to bring out the best of human creativity and ingenuity? So we'll start seeing a lot of new jobs being created in the workforce that is all about, for example, how do I train algorithms with better human intuition so that it starts mimicking that intuition in decision making? So if you look at -- one of the things we've done, we've seen in banking institutions, is compliance is a very complex process. And we did this work with a large financial institution where we look at IT assets and the user access patterns of how the users are interacting and accessing these IT assets, and then applied natural language processing to understand internal and external policies to start looking at compliance hot spots. Now you give that and serve these as insights to not just the compliance checker, who is looking for violations and threats in terms of compliance, but also give that visibility to the user and the IT departments to go make their jobs a lot better. Now the impact of something like this is you are now fundamentally transforming the productivity equation across the entire enterprise. You're using data to empower better decision making, you're using machine learning to remove the mundane tasks of actually reading documents and understanding that, to actually giving them insights and what they should be looking for. You still aren't removing the human from the loop. The human is very much in the loop. You're just empowering them to make better decisions, you're empowering them to bring out the best of what they do, which is bring out their creativity, bring out their ingenuity in the process. [What types of use cases] [should financial services organizations] [consider prioritizing?] There are a number of places you can start, but like I mentioned before, it's important to align business outcomes and start in places where you'll have a strong business impact and you have the data and expertise to execute. One way to start is the front and middle office. Traditionally, institutions have a view of their customers through a traditional CRM system, transaction history and portfolio holdings, but today, customers live their lives way outside these traditional boxes. They declare certain preferences to you, but you can also observe behaviors and how they interact with your products and services and you can infer or predict what their most preferred way, based on social, carrier data and so forth. At cognitive scale, we call this the profiler I. It's using AI to understand the customer at a very intimate level, a complete 360-degree view, and have this knowledge of their declared, observed and inferred characteristics. This could be for an institutional client, as well. If you're an asset management firm looking to launch a new fund or a product you can use this AI-powered understanding of all your clients to target the ones that are most likely to buy. If you have this information and you're an investment advisor, you can predict how a particular client will react to market movements or new sentiments and give them a more proactive personalized service. You can also use this 360 view for a client to predict life events and upsell, cross sell new financial products, be it an educational loan or a mortgage or give them better service. You can use this knowledge to define and build new financial products that can open new markets for you. I'd say start with the understanding of the customers. That's one of the easiest places to start. [What advice do you have for other business leaders] [looking to implement AI?] >> AI is an opportunity where the organizations that take advantage of it, sooner rather than later, will win and win against competition. But it's also very confusing, and there is a lot of noise in the system. And it's important to have a framework to get started with AI. First, understand the power of this technology and commit to being a business that will invest in AI. Now this is critical, and it's not just about experimenting with machine learning, but its corralling resources across the organization, getting the leadership to buy in, organizing for success, understanding the cultural implications, and so forth. Second, start with the business outcomes that you want to drive with AI. Too many projects fail when it starts with, "I have a lot of data, I want to do something with it, let's do AI." Understanding the strategic levers and the business outcomes that you're trying to drive is very critical to drive success. Third, identify and prioritize the use cases. I've seen very successful AI-first organizations use a matrix. It prioritizes use cases with maximum business benefit, by their ability to execute based on data expertise, and that helps them prioritize and come up with a list that they can start executing. Fourth, develop an AI center of competence and a road map to execute. You will need a track for constant experimentation, a track to move successful pilots to production in a repeatable fashion, and a track to train and scale and better your internal expertise. And lastly, I'd like to say AI is not a technology. It's a mindset of problem-solving, scientific method, dealing with uncertainty and constant learning and improvement.