4 steps to a successful AI implementation by top banking professionals

2021.07.07

Daniel Lee

PERSPECTIVES

The rate of emerging technologies and news of organisations implementing AI can often pressure leaders into implementing AI for the sake of it. While it is crucial for organisations to keep up with the maturity of AI, diving head in without careful consideration and planning can be costly. If you’ve suffered a burn as a result, don’t worry as you’re not the only one. In fact, almost 50% of AI projects often fail according to the Wall Street Journal.

 

1. Have the end in mind

As with everything we do in life, we ought to act with purpose and with a goal in mind. It’s the same with AI implementation. An analogy given by a senior banking professional from Emirates NBD was that of building a house. Before engaging a contractor to start digging the ground and laying the foundation, you would call in an architect and interior designer to plan how you want the house to turn out.

Are you looking to reimagine the customer journey? Drive revenue growth? Lower operational costs? Once you know the end goal of your AI implementation and how you will measure its success, the next step is to streamline your end-to-end processes.

 

2. Lean your processes

Automation is meant to enhance and simplify processes. If a process can be simplified without automation, do that first. This not only lowers the cost of automation but also prevents a case where you end up having a more complex process instead of a simplified one. As Sanil Mathews, Head of Operational Excellence at the Commercial Bank of Dubai said, “the first step is to look at the end-to-end process and the opportunities for re-engineering. Don’t automate a bad process.”.

 

3. Prepare your data sets

Preparation of data – maybe the most complex step yet fundamental since AI is fuelled by data. “For a successful AI implementation, financial institutions need to have that big data,” said Amit Malhotra, General Manager, Personal Banking from Commercial Bank of Dubai. Not having data is akin to driving a car without a dashboard. You don’t know what speed you’re going at and you don’t know how much petrol you have left.

At this stage, two main AI tools are often at play – robotic processing automation (RPA) and intelligent automation (IA). However, there is a common misconception that RPA and IA are one and the same. While both have the ability to extract data from documents, they work well for different types of documents. Sudhanshu Sawlani, Head of Robotics and Intelligent Automation at ING clarified this. “When you have structured data, RPA is at your rescue but if it goes beyond that, then you need some intelligence. You cannot just use RPA. You need intelligent automation to process the unstructured data.”

The ability to extract information from unstructured documents is critical as 70-80% of data is unstructured. Andrius Aneslauskas, Business Development Director at Anglo-Gulf Trade Bank shared his experience with TAIGER’s Omnitive Extract solution. “The key value-add from TAIGER was in reading unstructured documents like Board Resolutions and Proof of Ownership documents. We read those documents, extracted the data and pre-filled the different fields of the digital onboarding journey so clients could have up to 50% of information filled up automatically. We estimated from our POC that we achieved over 40% of time saved leveraging the solution.”

 

4. Make sense of your data

Ultimately, drawing insights about your customers and your business is what will help your organisation to grow. Success begets success. Achieving the objectives you set out for the implementation would give decision makers the confidence to further commit to future AI projects, and that is half the battle won.

 

Conclusion

Varun Bhatia, Head of Managed Services at KPMG Lower Gulf, mentioned that harnessing data and leveraging insights is extremely important. However, this often requires a combination of different tools. He added that while banks are progressing in the right direction, “what’s missing is the acceleration required to achieve the desired end-state”. From his experience, leveraging service providers who offer a combination of tools and resources has worked well for both banking and non-banking financial institutions.

Keen to kickstart another AI implementation? Check out the fulfilment and digital transformation services that our preferred partner, Electronics Document Centre (EDC) has to offer.

 

*Quotes in this article were taken from our #UnlockBetter Banking with AI Conference held during the DIFC Innovation Month. The conference was organised by TAIGER for industry experts to discuss how they’ve been leveraging AI to unlock “better” for their organisation. This means better in any form or structure, be it better customer experience, better efficiency, better ROI, and so on.

Keen to unlock ‘better’ with AI? Get in touch with us today.

Sorry

Your browser is unable to display our site correctly.
It is best viewed on a modern browser, such as Chrome, Safari or Firefox.