This article discusses the needs and challenges of enterprise AI for product innovation based on Winnie Cheng’s presentation at Data Science Salon NYC on the same topic.
Winnie Cheng is a director of AI in PWC’s innovation hub with more than 20 years of experience in artificial intelligence (AI) and machine learning technology. Her main focus concerns applying AI to finance, and she is co-founder of fintech startup, Flowcast.
According to Winnie, “AI and technology are very promising ways to tackle many of the challenges we face today, especially in finance”. Indeed, business insider predicts that banks can save $447 billion in costs by 2023 with the help of AI.
“Product innovation is an important way of unleashing the power of AI”, says Winnie. Harvard Business Review defines product innovation as creating new products or improving existing ones to satisfy customers' needs”.
Developing new products is what people usually understand as innovation. However, it is much broader than that, as it fundamentally involves solving customer problems.
It means innovation might give rise to an entirely new market or introduce new features to an existing product that enhance customer experience.
“Enterprise AI is the application of artificial intelligence and machine learning in everyday business processes to automate tasks, improve data-driven decision-making, and produce better results”, says Winnie.
Enterprise AI aims to find more profound insights into customers' needs. AI techniques are used to recognize trends in large amounts of data, so organizations can better understand the gaps in the market and find ways to improve customer experience.
Through enterprise AI, organizations can develop a targeted strategy toward providing a new product or new feature of an existing product that directly fills the market gap. Otherwise, it may be very costly to identify them through traditional methods such as focus groups, surveys, questionnaires, etc.
Also, with AI's predictive power, organizations can proactively discover market trends that may call for new product features.
In short, enterprise AI supports companies in discovering opportunities, deciding on which features to prioritize, and differentiating their products from their competitors.
“Finance is core to the well-being of society,” says Winnie, and adds: “without finance, people would not be able to afford and finance homes, and companies would not be able to grow and expand as they can today. Finance allows for the efficient allocation of capital resources, which is key to economic stability and societal advancement.”
Winnie also believes that finance is a data-rich industry. Financial institutions such as banks, hedge funds, insurance providers, and stock markets possess enormous amounts of financial data that flows at an unprecedented rate.
There are several use cases where product innovation can help improve the services that financial institutions provide to their customers. For example, leveraging data and AI can help companies forecast or predict behavior. This is especially useful in the case of mortgage loans, where banks need to determine the default risk of potential borrowers who want to buy a house.
Winnie summarizes the advantages of AI in finance as follows: “finance involves borrowing and lending, investing, raising capital, and selling and trading securities. All with many decisions that need to be made based on data and data analysis. Such analysis is the strength of AI and machine learning algorithms.”
We’ve seen above that enterprise AI can help financial institutions improve their products and services. However, product innovation with AI comes with some challenges that companies have to face.
Product innovation usually requires companies to collect and analyze relevant data. Enterprise AI can help with such processes. The main challenge lies in dealing with the sensitive nature of the financial industry and firms operating under strict regulations. Such regulations may restrict the collection and use of personal data.
“Organizations must use AI with care, so they can continue to demonstrate that in using AI machine learning algorithms within their current decision-making process, they're still able to uphold their legal obligations and avoid any conflicts of interest,” states Winnie.
Another challenge is to identify unintended bias in the AI model. For instance, banks may use AI models to assess a client's creditworthiness, and “a single declined loan application can change a person's life,” says Winnie. Therefore, organizations should select their models with extra care.
Data security is another challenge that needs to be taken into consideration. “We must ensure that the AI feature does not introduce additional risks by exposing data during the training or inference process”, highlights Winnie.
Are you interested in joining the AI conversation with industry leaders and senior practitioners? Check out the Data Science Salon events 2023, early bird rates are now available.