How AI is helping financial institutions automate and accelerate | data, datarobot, automation, artificial intelligence, AI
Artificial intelligence (AI) is rapidly transforming the operations of financial institutions, making them more efficient and competitive.
AI can help accelerate revenue growth, reduce risk, and lower costs, delivering tremendous value to financial companies. Its applications in finance are numerous, said Yeo Hwee Theng, data scientist at DataRobot, in a webinar co-hosted with Finance Asia in June.
In a fully AI-driven organization, AI does the heavy lifting, allowing financial institutions to focus on their essential work to better serve customers, said Nina Xing, another data scientist at DataRobot, in the same webinar.
In the case of anti-money laundering (AML) practices, AI can help financial institutions save substantial capital. The estimated return on investment can vary between $ 2 million and $ 5 million, according to DataRobot studies.
AML is important for reducing regulatory risks and increasing brand awareness. However, the rule-based models deployed by most banks are not the most ideal AML solutions. Machine learning models can be introduced to replace or supplement existing rule-based models.
Businesses can create rule-based models throughout the transaction monitoring process in their AML practices. The models collect information about customers and check if they are legitimate when onboarding customers, Yeo said.
The banks then monitor the transactions. Some of them will be flagged as suspicious by rule-based models after setting up alerts.
“But the problem with rule-based models is that they are rigid and slow to respond to changes in customer behavior. And therefore, many alerts are in fact false alarms. And no suspicious activity report will be filed after the investigation, ”she said.
Only around 40% of alerts concern complex situations. Studies have also shown that forgery rates can exceed 90% with rule-based models and that huge resources are wasted during the manual investigation process, she said.
“[In comparison], machine learning models can filter out false alarms and reduce the number of cases requiring manual review. Thus, they increase operational efficiency and reduce costs, ”she said.
When it comes to ensuring that models are always up to date and relevant, Yeo added that AI models can be recycled periodically with the latest transaction data to capture changes in the state of the market and customer behavior. The recycling process, which is fairly straightforward, can be scheduled to run on a regular basis.
AML only demonstrates one of the countless applications of AI in finance. On the buy side of the business, AI can be leveraged in areas such as asset allocation, building factor models, and discovering smart beta strategies, Yeo said.
Sentiment analysis is another good example of how natural language processing, which is part of machine learning techniques, helps drive business. It can help capture signals that traditional quantum techniques struggle to do, she said.
It’s important for financial institutions to embark on the AI journey as early as possible, which begins when they establish awareness and acceptance of AI across the organization.
As they move up the AI maturity curve, they will develop more use cases, improving the efficiency and standardization of the entire pipeline from end to end, while accelerating their data science capabilities.
When many use cases are generated, organizations will begin to “democratize” data scientists, allowing staff and business analysts to work on AI initiatives. These are the key steps on their journey to become fully AI-driven, Xing said.
However, automation is not easy to achieve, not least because people have to trust the model they have built in many cases.
“People can be skeptical about AI and think they are offering black box solutions. Others want to ensure that the models remain suitable for constantly changing business environments and comply with global regulators, ”she said.
The technical details present other challenges. Financial institutions need to know the accuracy of the model, compare performance between models, and understand the trade-off between different aspects such as speed and accuracy.
They should also be able to explain the structure of the model and understand the type of data needed for the preprocessing step. Measures must be put in place to avoid data drift.
“All of these questions are important to validate that your AI is not a black box. And we don’t rely on unexplained models to make critical business decisions, ”she said.
These are difficult but important steps to take in order to reap the ultimate benefits of AI and stay ahead in a rapidly changing business environment.
To learn more, watch the on-demand webinar here.
Haymarket Media Limited. All rights reserved.