Skip to Content
17:21 Video

Accelerate Session - Innovating with AI in the Financial Industry

Innovating with AI in the Financial Industry: 6 Steps for Turning AI into ROI
Click to View Transcript
Mm Hello everyone. Welcome to my accelerate session innovating with AI and financial services. I'm Diane. Associate financial services director here of pure storage and I focus on our solutions and financial services. Fintech and Read tech.
Thanks for joining me for this session where we're going to be talking about practical AI and how to turn your a into ry from market insights that chatbots fraud detection to algo trading. Ai is driving innovation and automation in financial services but with challenges around compliance, legacy technology and silo data but firms need is a practical approach to implementing AI across the organization.
In this session, we'll look at the foundation needed for a successful enterprise AI and six steps for turning AI into our way. Finally, we'll try to pull it all together and we'll touch on how pure can help you accelerate your AI initiatives. Until recently, the promise of artificial intelligence or AI and machine learning was
more the province of science fiction than real world application. That's changed dramatically over the past few years. But we've really only begun to scratch the surface of opportunities in practical AI and machine learning as investment and opportunity here is expected to grow dramatically in the years ahead. The current investment has the potential to
generate outsized returns down the road. Given all of this, I think it's safe to say that artificial intelligence and machine learning have officially transitioned from cool to vital where previously most organizations used aI and ml only for peripheral functions such as chatbots. They are now key to mainstream productivity and revenue drivers especially as financial
services becomes faster and more global and much more digital advances in compute and storage resources, new and abundant data sources and flexible development tools have all enabled financial institutions to leverage AI to handle sensitive financial transactions, an end to end customer activities such as investing and lending and critical functions such as fraud detection.
In fact, McKinsey estimates that artificial intelligence can generate up to $1 trillion Ahead, An autonomous research predicts that by 2030 Ai will allow financial institutions to reduce operational costs by 22%. Some of you may be working directly with artificial intelligence today and some of you may have no interest at all in A I for its own sake.
But you're very interested in the use cases and the business outcomes that yeah, I can help drive no doubt AI is having an impact on nearly every facet of business in financial services and in other industries. But several extreme examples stand out as having particular promise for delivering the greatest impact in financial services and in many cases in other industries.
These include trading and investment fraud detection and prevention. Smarter and faster decision making, predictive analytics, client risk profiles, regulatory compliance, including know your customer and AML and customer experience. I'm sure most of you are inundated with stories, heralding the benefits of AI and its close cousins, machine learning, deep learning and natural language processing and those benefits
are real In fact, a recent and video survey of financial services professionals showed that 83% of respondents Believe it is important to the company's future success and 34% said Ai will increase their company's annual revenue by at least 20%. But achieving enterprise level success and roi requires the right foundation. So let's take a look at the six steps for turning your Ai into our Oi,
whether you're looking to implement AI for fraud prevention or better customer insights or to improve efficiency with hyper automation, something that Gardner identified as an AI megatrend. The test will come in moving from prototype to measurable roi to do that. The first and arguably most important step is to do A I with a purpose.
One of the most obvious wise for automation is eliminating low value tasks from workflows so that valuable resources can focus on more strategic work and wealth management. For instance, automation can streamline account setup and client on boarding including KY C. Requirements and thereby improve customer experience, reduce errors and freestyle from box ticking, allowing them to focus on more valuable work.
Another important y could be using AI and machine learning to quickly identify fraud patterns, weed out false positives and block malicious activity before it impacts the business. An effective machine learning strategy can automatically detect hidden fraud by focusing on subtle pattern changes and unlike traditional rules based processes, it actually enables algorithms to become more efficient and
effective as datasets increase once you understand your why you'll also need to be able to communicate it to your stakeholders? The long term success of your AI projects will require the ability to move beyond unrealistic short term financial gains to a broader understanding of AI is role in transforming a firm and enabling its longer term strategic goals.
Without setting expectations. At the outset, stakeholders may have overly optimistic expectations for AI as a plug and play solution as opposed to an iterative and new way of working. Okay, so you have the why now what's the what what is the business problem that you're looking to solve and what outcomes can you expect first?
You need to look at an algorithm feasibility. Is this a problem that lends itself to an hour ago? Not everything does. Of course. Either because of the type or quantity of data needed or the nature of the events to be analyzed. A I can't solve all problems here. You should also look at whether there are
existing out of those that fit your problem. If not you need to factor in time to build, test and tune out goes to address it. Of course, you also need to understand the potential impact of the project. Well solving this problem provide valuable impact on the business with the end result. Have tangible business impact on revenue, costs, customer expectations,
risk and a related question is recurrence. Will your a I project address a recurring issue or is this a one off situation? How often does your current challenge arise? And how often will this solution be utilized? And lastly of course, you need to look at the data. What are the data challenges with the Ai solution you're looking to implement?
Do you have and can you access the right type and quantity of data to enable the model to be trained to be accurate? So now that you have buy in and budget from the C suite and you've determined you can access the data. You need it's time to get your data house in order. In a recent expert io survey of AI articles.
Not surprisingly, the most common topic was data. And by 2024 Gartner predicts that over half of all finance organizations will encounter problems scaling with AI solutions the underlying infrastructure to move your A. I. Project from a great idea to improving business outcomes must be able to handle massive amounts of data without it,
your team may not be able to provide these apps with the quantity or quality of data they need to generate. Ry specifically, your data storage will need to be able to support the simplification and consolidation of unstructured data. Well, most processes process automation up until now has focused on the simpler to use structured data. Unstructured data is exploding along with the
digital economy. In fact, Gartner estimates that unstructured data represents an incredible 80-90% of all new enterprise data And it's growing three times faster than structured data. So what is needed is an infrastructure that addresses three key challenges. 1st, consolidation of unstructured data, whether you're leveraging texts and images for
market analysis or earnings transcripts and financial filings to generate complex trading algorithms, you need storage that can seamlessly handle this often unwieldy unstructured data. Second performance demands, performance is critical across the entire AI and analytics. Spectrum machine learning and software development.
Workflows require fast storage and massive throughput And 3rd data reuse the analytics. AI continuum creates a need to reuse data across applications and even to utilize new data that's created by your AI and Ml applications. Enterprises looking to successfully deploy AI must ensure that their data infrastructure is up to the task addressing those challenges.
You're created Flash Blade as a unified fast file and object storage platform designed from the ground up to meet the demands of modern data. Flash Blade can help with consolidation of unstructured data and his throughput hungry applications demand more the massive, massively parallel architecture of Flash Blade is perfectly suited to address these applications needs and his data reuse grows.
Flash Blade can serve as a central repository rather than copying data to all potential applications. Enterprises can also leverage storage and compute in one platform engineered specifically for AI projects. Solutions such as Harry and flash stack for AI offer high performance. Architecturally optimized solutions that can harmoniously run within existing data centers
and can manage any workload on any mode node at any time. Okay, let's talk about building a winning team. It might seem like hire more data scientists is the key step here, but the talent to run AI initiatives is only a fraction of the equation to success. Yes, you'll need the right people, but also the right mix of people.
If your data scientists are busy acting as data herders, your team can't be efficient. The human resource element is every bit as important to success as the physical handling of the data itself. Organizations with the greatest effectiveness in making Ai integral part of their business strategy. Use mixed role teams and leverage an Ai ops or ml ops approach that delivers a flatter more
agile workflow arrangement that emphasizes teamwork rather than vertical handoffs related repeated iterations, test train tune, rinse, repeat are needed to arrive at usable systems so the team needs to be a near constant contact to experiment and learn in a series of short repeatable steps. Successful organizations have also found that aligning Ai to business initiatives with
stakeholders engaged throughout is the best way to deliver value. Both the diversity of the team and leveraging their perspectives and insights for the project are key. This diagram I think originally produced by google is an oldie but a goodie that I think illustrates what we've been talking about AI and machine learning code is central, but only a very small piece of an overall Ai.
Machine learning lifecycle. Number five nailed down the feedback loop. This one I think is often overlooked. But AI works best when it can learn from itself to improve and stay on top of measuring and reporting will help with continuous improvement knowledge gained through effective AI is also multiplication while the initial gains may be
small if done correctly, they can be expected to grow exponentially over time. So think about how you will quantify the the R. O. I. Your wife will dictate the KPI s here. Is it time saved to free resources for more productive activities? Is it mistakes avoided that reduce risk or is it new trading opportunities that provide additional revenue?
Gardner also notes another benefit of measurement in terms of identifying opportunities for growth and scale of AI from Gartner. Embracing metrics enables organizations to showcase how AI can be used across the enterprise by highlighting its benefits and risks in certain areas. For example, the ability to analyze video or images might start in the security realm,
but with some maturity could be used to analyze its organizations, foreign presence or understand how customers react to products. This also ties back to our last point, ensuring that your team is broad and diverse as they may. Uh diversity of thought may introduce new ideas for the AI that you're utilizing. Lastly continuously auditing data science processes and outputs including the quality and
relevance of the input data can help identify model drift and ensure reproducibility. Both of these are are important factors when looking to implement explainable AI which is becoming increasingly important to address questions of fairness and bias, especially in financial services. Okay, this last one maybe a bit obvious, but unfortunately there's a little bit of magical
thinking attached to A. I. And business leaders may expect near instant results from even initial forays into AI capabilities. This may be partly due to hype from vendors and the media and partly due to an over reliance on technology to answer all challenges, but nothing will kill A I. Efforts faster than unrealistic expectations.
In reality Ai success is a journey not a destination, so patience is needed. My young padawan while this can be mitigated by focusing on low hanging fruit opportunities for early efforts. As we discussed earlier, it's important for the treat team and the stakeholders to know what is and isn't realistic from the outset and to keep everyone in the loop. It's not an overstatement to
say that AI is fast becoming table stakes enterprises need to adapt now or fail in the future knowledge gained through effective utilization of machine learning and AI is multiplication so that if done correctly it can be expected to grow exponentially over time. Conversely, firms that fail to invest may find themselves hopelessly behind before they know it threatening their very existence at the same time, it's also important to take the time to
do things right. This means understanding not only the business challenges but also the optimal technology for each task and how to deploy it. And finally it's important to remember that AI is not an I. T. Or data science exercise. You're a I. Initiative won't bear fruit without the active
and sustained executive support. Before we wrap up, I wanted to take a minute to look at one of our financial services customers that successfully leverage pure to advance their AI and analytics objectives. Liquid net is a global institutional investment network providing liquidity and investment opportunities to the world's largest asset managers.
Liquid net historically collected data mainly for post trade analytics. The current investment environment is more complex and faster paced and they now need to provide traders with more real time analytics that will allow them to make pre trade decisions. This wasn't feasible previously because legacy storage systems lack the IO and parallelism to move very large data sets fast enough
liquid nuts, application development and analytics teams employ modern tools like elasticsearch spark and Kafka and they need infrastructure to keep up with workload demands, literally processing millions of events per second. In addition, their teams regularly pursue several projects concurrently. Flash blade gives liquid net the performance flexibility and scalability.
They need to support multiple disparate workloads concurrently and as a bonus Evergreen allows them to keep the same footprint while adding capacity and benefiting from the latest technology as their AI projects grow. I hope this session has sparked your interest in practical AI in ways that pure can help accelerate your AI initiatives for more information appears smarter and simpler
infrastructure for AI and analytics. And to find out why Harry was recognized as the best AI solution for big data and meta chose to partner with bureau for its AI Research Supercluster. I hope you'll visit the financial services page on pure storage dot com and check out our financial services blogs or even better. Let's talk. Thank you for joining me for this session on
innovating with AI and Financial services. I hope you enjoy the rest of your accelerate tech fest 2022.
  • Artificial Intelligence
  • Video
  • Financial Services
  • Pure//Accelerate

From market insights to customer experience, fraud detection to algo trading, AI is driving innovation and automation in financial services. But with challenges around compliance, legacy technology and siloed data, what is needed is a practical approach to implementing AI across the organisation.

Test Drive Modern Storage for Analytics and AI

Test Drive Modern Storage for Analytics and AI

No hardware, no set up, no cost–no problem. Take a free test drive and see how easy it is to scale every tool from ingest to visualization.

Try Now
Pure Storage FlashBlade for Electronic Design Automation (EDA)
Test fast and iterate better for an accelerated EDA production workflow with improved chip design quality enabled by Pure Storage.
Solution Brief
3 pages
Continue Watching
We hope you found this preview valuable. To continue watching this video please provide your information below.
This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.
Your Browser Is No Longer Supported!

Older browsers often represent security risks. In order to deliver the best possible experience when using our site, please update to any of these latest browsers.