Sathish Muthukrishnan, Chief Information, Data and Digital Officer at Ally Financial, contends that everyone should be a part of the AI implementation, including generative AI enhancements, and it should be an enterprise-wide transformation initiative. In this keynote presentation, he will address this position and surrounding questions, specifically calling on his role in leading technology at a Fortune 500 financial services company and his deep background in leading digital transformation. This discussion will dive into the components of AI Transformation at the enterprise level – a journey full of many unknowns, great passion and the need for both technical proficiency and risk management.
The transformation journey will take everyone in an organization, from top leadership to customer-facing team members, all of whom are enabled in their roles by the digital expertise found in the technology group.
What you'll learn
- How to determine if your organization should explore AI and articulate creating and capturing value from this investment
- Who should be engaged at the Board and leadership levels of your organization, and how to generate buy-in
- The importance of partnering closely with control groups like Compliance, Risk, Audit and Legal
- Why you should develop an enterprise-wide AI playbook and a governance structure that sticks
- Ways to include AI development and governance into your established processes
- How to generate excitement among employees and give them the tools to learn (with guardrails)
- Why it's critical to engage with external stakeholders, including regulators, shareholders and the media
Transcription:
Sathish Muthukrishnan (00:10):
Good Morning. It could be better. For decades, we have seen AI being negatively depicted in sci-fi movies as that mean robot that takes over the world or some Hi-Fi spaceships that battled out using lasers For decades, technology has been a cost vacuum or a support beam at best. The time has come. The time has come to change this narratives, to break the barriers, to stop the madness. We are amongst the greatest innovation in technology that perhaps will allow us to dream a little bolder. I have a dream for the upcoming decades. I want to change this narrative. I want technology to be a value creator. I want technology to be a revenue generator for all our banks and companies. In fact, I suspect this is part of every technologist's dream.
(01:30):
If we safely and securely scale AI, maybe we can realize this dream. In fact, my dream is to generate millions for ally by scaling AI safely and securely. Let me be candid. I am living the dream right now. My team is living the dream right now. So today I want to share with you why we embarked on this AI journey. Why did we dream bigger? What did we do to bring our entire company along on, and also share some lessons that we have gotten this journey to help you in yours. So let's start. Let's start where all good business decisions begin. The why should we embark on an AI journey as a financial institution? We may not be battling this space robots, the mean robots that we see in sci-fi movies, but we embed and experience AI in our personal lives. Think about that intriguing movie that is being offered up to you while you're watching Netflix. Think about YouTube meticulously learning from your mouse clicks and keyboard strokes to give up that next video that will keep you engaged in the platform. Think about the next song that Spotify gives you so you can continue to be engaged in that platform.
(03:15):
So whether it is ordering a meal, tracking or keeping track of your health habits, AI is embedded in your personal lives every day, in every action. Perhaps some of you are wondering if Gene AI wrote the speech of bind, guess you'll never know. So scene two, why is it AI not having a profound impact in our professional lives as it does in our personal lives? The question is, can we bridge the divide between our professional and personal lives using AI? That question is not just in a technology leader's mind, it is in your customer's mind. The customers are asking, why can't my bank provide the same intuitive, personal, delightful experience that I face in my life every day? That is why we embarked on the AI journey. It is the right thing to do for our customers. us. In fact, these three words do it right, drives our thoughts, our strategy and actions At Ally, it is fundamental to our culture. It is fundamental to our actions. These three words pushed us to start a digital bank right after the worst financial crisis this country has ever seen. These three words pushed us to evaluate, underwrite, and recast your auto loans digitally during COVID for our customers, these three words pushed us to be the first bank to eliminate overdraft fees for our customers.
(05:18):
These three words also drive our technology investments. Having seen our AI journey, people think that we have a crystal ball. No, it was those three simple words guiding our technology investments for us to build a software defined network that has contemporary security controls and privacy controls. It is these three words that allow to centralize enterprise data in a single data warehouse so it can be easily accessed and used for compute. It is the right investment that has allowed us to have a cloud infrastructure for our applications to run and be expanded on scale and be able to access these data easily. And because of that, we are now in a great position to take advantage of this technology revolution. But is that enough? You need to get the buy-in from the larger population across your company whenever a new technology is introduced, if people don't understand it, they fear it. And I can totally understand that If I was not involved in this technology day out, I would be apprehensive too.
(06:40):
I believe it's the technology leader's responsibility to eliminate that fear, and I'm going to share with you how we did that. It started from the top. I vividly remember having a conversation with my CEO. I asked, everybody's talking about this Gen AI, should we experiment and forge ahead or should we wait and see how our competitors are using it and we should wait to see how it evolves? His answer was, and I coach, my job as a CEO is to balance risk with innovation, but it is your job as a technology leader to push the organization to innovate. That hit me hard. That 10 minute conversation led me to a series of ideas, thoughts, questions and promises. So following that, I scheduled a deep dive with my CEO. I realized it was important for me to showcase the power of the technology, but more importantly, to eliminate fear, show how we are protecting it and how we are protecting from it. So we did the deep dive. I wanted to show the power of the technology as it exists today so he can reimagine the promises it can bring to our businesses tomorrow. I walked away from the meeting feeling a ton of support to move ahead and to innovate. So I did what every technology leader does to innovate. I brought in all of the control and governance groups.
(08:26):
It was important for me to get their support to get their buy-in.
(08:33):
I vividly remember again having a conversation with my chief risk officer saying, I will not have answers to all the risk questions that you're going to throw at me, but I promise you we will figure it out. I promise we will balance risk, the rewards of innovation that all of us can enjoy together. A series of conversations with these control groups gave me clarity that I needed to centralize AI execution. I needed to centralize the execution of these controls. It was important for me to showcase how we are protecting the technology and how we are protecting ourselves from that technology. Hence was born Ally.AI. Ally AI is a single platform that serves us about between Ally and all of the available commercial LLMS out there. Ally AI may not be ready to take on that mean robot yet, but it does a number of great things.
(09:41):
The first thing it does is to inspect the input from Ally and the output from the LLMs. It has in real time the ability to inspect, identify, and remove PII before sending the data outside. It also has the ability to rehydrate that PII. So the consumer that is using the ally AI doesn't lose the context. Very powerful. This layer of security that we built started to build comfort with the C levels, my CEO, my risk leaders so much so in one of the board meetings, my chief risk officer goes, the strategic risk of not using generative AI is greater than the operational risk of using it. Let me let that sink in. This is the chief risk officer that is responsible for setting the risk appetite for the entire company. Articulating that the risk of not using generative AI is higher for the company.
(10:50):
We had won half the battle. We now have the founding Technology foundation built. We now have the support of the executives. It was time to move. It was time to execute. But let me tell you, figuring out this complex technology in this banking regulatory environment was not easy. So we had to go back to our founding principles of finding a simple strategy with guiding principles. So we had three non-negotiable principles. The first one is that technology is fast evolving. Let us focus on internal facing use cases that faces off to our own employees. Number two, have a human in the middle with the permission to intervene and change the output coming from the generative AI large language models. And number three, protect personally identifiable information and not allow any of the large language models to learn from the ally data.
(12:05):
So with the simple strategy with the tech foundation in place, with the support from our control partners, with support from c-level executives and a culture of innovation, our dream was starting to become a reality. Now I want to share how some of our colleagues are living the dream. Our first use case was based in our customer care associate. Our call centers today, they take about 10,000 calls. The call handling time is between 14 and 15 minutes, and at the end of every call, they painfully think through what the conversation was and capture the summary for regulatory and consistency reasons. Ally AI in real time transcribes that entire conversation creates a summary and presents it to our customer care associates. For them to either change it or accept assets. The accuracy has gone up 60%. The acceptance rate rate is at 90%. It has singularly allowed our customer care associates to focus on our customers and not worry about multitasking while they're conversing.
(13:19):
Our marketing and user experience teams provide content on a daily basis. You might be surprised for a hundred percent digital organization, this starts with a blank paper. This is content that you see on our websites text that you see when you do a dropdown on a menu content that you see educating our customers to be responsible with their financial needs. You get the idea Ally. AI now creates the first draft at a pace 60% faster. It has removed writer's block now. It's allowing our employees to focus on things that they never would've gotten to or improve upon the first draft that they get from generative AI. Our investor relation teams, small but mighty. Every time earnings period comes around, they have to look at 25 of our peers read through the earnings report transcripts, the analyst asking the questions, what is the CEO saying? And start preparing Ally for our own earnings release. It used to take them four to six hours to read through every report. Today, ally AI sends them an email with a summary of what each of them are reporting, what questions are analysts curious about and what topics were discussed. The four to six hours has become four to six minutes. So now they can focus and dive deep on things that are important for Ally.
(14:50):
Even our control teams can use a jolt from generative AI. Take our audit team for example. They're responsible for auditing the capabilities across the entire company. They may not work in technology, but they have to understand the tech capabilities. They have to understand all the test conditions that are required to audit this capability. They have to create a risk matrix and at the end of the audit, they have to create a report. Ally AI does this all for them so they don't have to focus on understanding what it is, but focus on auditing us to make us even better. All of these examples has created a bus around the company. Everybody wants to participate in this transformation. Everybody sees the art of the possible. We have over 450 use cases in the pipeline for generative ai, and most of these use cases are coming from outside the technology team. But you see the challenge. I can't execute all of those use cases at the same time. We need to be able to prioritize and focus on what we can execute today. This brings me to the lessons that we have learned and I want to share with you. Number one, deliver value
(16:16):
For this. You cannot be a technology leader. You have to think like a business leader.
(16:24):
You think like a business leader, you would remove the emotions of the technology and focus on what use case is going to generate the most money is going to drive the best productivity is going to drive the best efficiency for our company. That allows you to prioritize the work effort and make sure you're getting the best return on investment for the time and effort spent on these use cases. So focus on that. Drive value Number two. This is not a technology transformation. This is an enterprise transformation. You need the buy-in and participation of everybody in your company. So what can you do to educate and empower your people? People have various roles, they have different skillsets and they have different levels of understanding. We did it through what we call an AI playbook. AI playbook. Imagine it as the Uber Google translator regardless of employees skillset, understanding, role persona, they can talk in their own language and playbook allows you to understand how you can take your concept and go to execution and use AI effectively. It obviously contains the non-negotiable principles, but more importantly, it has the governance structure around how your idea can be progressed to execution.
(18:00):
We bring AI Playbook to life through what we call AI days.
(18:09):
AI days happen once every four to six weeks, and it's a four hour time block for the entire company. Anybody across the company is invited to participate. It has three components. One, we have external speakers talking about the advancements in Gen AI to inspire us. It has the internal consumers of the technology capabilities that we roll out to talk about their struggles and successes. And number three, technology talks about what have we advanced within Ally for others to participate in. We have seen about 1200 to 1300 participants for every AI day. Again, majority of them coming from outside of technology. We even asked some of our employees how do they feel now that AI is part of their job compared to when they initially heard Gen AI is going to take over theirs? Let me roll the video.
Video Presentation 1 (19:12):
We're in the business of relentlessly caring for our customers. And Ally AI is a game changer in helping us create content that's relevant to their financial goals almost instantly whenever we need to, wherever our customers need us. So on my team, we're using AI to develop first drafts faster. Things like language for our website, our mobile app, other communications that are really important for our customers. That frees us up to focus on doing what only we can do, which is understanding and caring for those customers. It's really exciting to have Ally AI assist us in that mission.
Video Presentation 2 (19:44):
AI summarization has really helped connect with my customers more with the genuine conversation where I can focus on that while AI is doing the summarization for me and taking that stress off of me.
Video Presentation 3 (19:55):
Ally AI call Summarization tool has saved me minutes across each and every call I have every day. Ally AI has improved my overall quality of being able to provide that world-class service to our customers, which is what I'm here for. I love the human interaction part of my job. It feels great to be a part of a company that's innovative and on top of the trends and not scared of change.
Video Presentation 4 (20:25):
Ally AI empowering employees to do more.
Sathish Muthukrishnan (20:31):
Ally AI is giving freedom to set the genesis, to focus on customers and not worry about keeping track of that 15 minute conversation to create a summary. Think about this. They are feeling proud about being part of an innovative culture and a company and being tech forward. Customer care associates may not be the first to learn new technology or adapt new technology, but they know your customers best if they are driving this transformation, if they are the voice of change, think about how powerful your AI transformation will be. That leads me to my last lesson. Be courageous. To get to be courageous, we have to fight through fear. Every time a new technology comes around, there is fear, there is anxiety, but then it becomes part of our life and it opens up. New industries shows new commercial opportunities. The biggest fear with gen AI is that it'll take away our jobs. So can you wait in the sidelines, wait to figure out how everybody's using and indulge in the fear? Maybe not because one of those spaceships will come and hit you. We have to eliminate fear. That is why the human in the middle to begin this transformation is so critical. AI can tell you your heart is beating, but can it tell you why?
(22:15):
You may be in love, you may be sad, you may be happy. You need a human to understand that complexity. You need humans that understand, appreciate, and indulge in this transformation to drive this forward. Humans driven with purpose through a simple strategy is what will power your transformation. So this notion of delivering value, educating and empowering your employees, and finally being courageous leads me to this flywheel. If you educate your employees and give them permission to experiment, they will feel empowered to be part of the change. When they do that, you eliminate fear and then you can move from experimentation to execution and start to realize your dream. So as I wrap up, I have some challenges for you. What are your non-negotiable principles that everybody can understand? How will you use AI to drive and deliver value for your company? What is going to be your AI playbook to educate and empower your employees to drive this transformation? And are you courageous? Are you courageous enough to let your customer care associates lead the change? Are you courageous enough to have the conversation with the CEO and the chief risk officer? These are the questions that you need to answer to make your AI transformation successful and make your dream come true. Thank you.