Basic understanding of artificial intelligence (AI) in finance?

Introduction

The topic of artificial intelligence (AI) is a curious one for all. Every finance person would read this cautiously to see if it is a threat or an opportunity. Before giving a straight answer, let me rewind to make you understand the history of computerization, and you will find the answer that if you embrace change faster, you will be ahead of others.? "Who moved my cheese ?" by Dr. Spencer Johnson is very apt to get your answer, too, in this context.?

Many in their 50s now would have seen the phase of computerization displacing the role of typists and stenographers. Initially, there were data entry operator roles, and today, we all are typing, writing minutes, and taking notes of meetings. Now, technology has gone to the next level, where digital assistants are added to your virtual meetings, and transcripts of your sessions are made available immediately after the meeting ends. The fundamental shift is that huge volumes are being managed. Had we continued with typists/stenographers, the scale would have been limited at this level. Then, we all feared job loss, but quickly, people adopt, enhanced their skills, and made the best of the opportunity ahead of them. Some laggards grew little in their career; this is what history tells us. So, technology change (AI) is knocking at your door, and a threat is visible shortly, so it's better to get ready to unlearn and relearn. Stay caught up; this article explains some use cases to enable you to relate and visualize changes around you, explain what new roles will emerge soon, and how one can embrace change by learning or acquiring new skills.

In today's world, AI innovation and adoption are evolving every passing day and growing without any boundaries. As the use cases become popular, the speed of adoption is increasing. Every corporation is now discussing having an AI strategy for their business to increase revenue, optimize cost, and run processes efficiently and effectively. Artificial Intelligence coupled with Business Analytics will be the new revolution mantra.

Why AI and What is AI??

• Data, often called the new oil, is constantly generated and stored globally. AI, particularly machine learning, is increasingly used to analyze this data and make informed decisions. Our decisions in finance often depend on data, whether it's structured or unstructured. In the case of structured data, we can build logic and automate processes. But when it comes to unstructured data, it's more complex and often requires human intervention. This is where AI comes in. It can analyze unstructured data and provide insights that can aid decision-making. The need to adapt and understand its role in decision-making becomes more urgent as AI becomes more prevalent. The usage of AI will be relevant and economically viable for substantial data volumes and repetitive activities when the speed of decision-making is equally important and when humans cannot make decisions at that pace.??

• Let me explain this with an example: In a traffic signal, we want the police to catch the violators of the signal undisputedly. If this must happen, the police officer has to sit in an elevated position opposite to every signal to spot the violation and to prove that you/your vehicle violated the rule. This is not doable in manual mode in flowing peak traffic, further many humans even if deployed cannot bring a fool proof result at scale. The next step is sending out challans to violators and seeking paying fines on time centrally, an administrative challenge.? What are the changes that are happening? Now, both speed cameras and traffic cameras are set up where the images of violators are captured along with number plates and violation evidence with proof that when the signal was red, the person crossed the signal. The ability to handle a massive volume of signals, challans with evidence showing violations to violators, tracking of payment of fines, and building repeated offenders' history is done at ease. Insurance companies can leverage this data to charge higher insurance premiums for consistent violators who are prone to accidents. This is the power of AI combined with Technology and Automation. The volume of image recognition and processing is humongous. This is just a narration of one use case, and across the world, there are many use cases that people are developing for adoption in multiple scenarios. You can use the same idea in your colony when visitors go in and out. Work can be managed with fewer security staff and can be 24x7, too.

To enable decision-making, machines are trained with structured or unstructured data to give us the desired output with progressively higher accuracy levels. More efforts are put in to improve the accuracy levels, and once Machines are trained, they learn on their own when exposed to the new set of data; they understand the patterns in the data and enable us to make decisions indicating the accuracy levels. When we train the machine, it is called Supervised learning, and when it learns independently, it is Unsupervised learning. Now, you can relate to the terms AI & ML. This is nothing but Artificial Intelligence and Machine Learning interplay.

Industries where AI is adopted and typical use cases:

Typically. the early adopters are the BFSI sector and the government.? The use cases where the adoption has happened are :

• Customized chatbots and virtual assistants.
• Fraud detections
• Investment decisions
• Insurance claims settlements
• Loan processing
• KYC?
• Predictive analytics to forecast sales and customer churn analysis
• Customer onboarding to give a better user experience
• Market research: sentiment analysis, customer behavior patterns

AI adoption and benefits one can expect:

AI adoption can very quickly trigger transformative benefits, as below:?

• operational efficiency,?
• customer experience,?
• risk management,?
• compliance and innovation.?

It can also create greater agility, profitability, and resilience in the evolving digital economy.

Some of the use cases in various categories:

• Fraud Detection

In fraud detection, we are witnessing increasing AI adoption. Key characteristics are:

• Real-time monitoring: pattern recognition & behavioral biometrics
• Adaptive learning capabilities—supervised and unsupervised learning
• Enhanced accuracy and reduction of false positives

For example, in the case of credit card spending, if there are any abnormal patterns in the spending— value, volume, location, etc —then a fraud alert trigger can happen. Depending on the case and the rules the respective companies would have set to tame the model, it can be a digital or human intervention.?

• Risk Management

In the risk management space, sometimes decision-making must be faster and quicker. AI comes in handy for deployment in a couple of these activities to expedite the process or bring sharpness in decision-making:

• We are leveraging advanced analytics, automation, and real-time monitoring capabilities.?
• Enhancing data-driven decision-making, automating routine tasks, and improving predictive accuracy
• Proactively identify and mitigate risks, optimize portfolio performance, comply with regulatory requirements, and enhance operational efficiency.?
• Driving innovation and resilience continually

Use cases for the use of AI is witnessed in credit assessments, credit risk, stress testing, assessing risk tolerance, etc.

• Adoption of AI-powered chatbots and virtual assistants to enhance customer service in banking in India and other industrial applications

Most banks have started deploying AI-powered chatbots and virtual assistants to enhance customer service. What is being offered are :

• 24/7 Availability: Round-the-clock and instant responses
• Personalized Assistance – recommendations and NLP (natural language processing) capabilities
• Transaction Support - Balance Inquiries and Transaction? alerts
• Customer Onboarding and Support – Account opening and product information
• Problem Resolution and Escalation – Issue resolution & seamless handoff to manual agents.
• Cost Efficiency and Scalability: Operational Efficiency & Scalability.

For example, BOTs, or digital assistants, help in banking in a big way. With our collaborating partners, we at CFO Bridge are enabling our clients with BOTs or digital assistants to work on AR reporting, push mailers, reminders, and information support. We have also built MIS and analytics dashboards as an extension.?

Some key advantages are that the BOTs will work odd hours and keep the work ready for the teams to take action, which will be managed by exception. The advantage would be that it will work in the background as if a human has delivered the work, greatly benefiting from handling a massive scale of work.?

An extension of these would be robotic process automation (RPA). And there are much deeper use cases of these, and innovation is happening rapidly in this space.?

Even Chat GPT is also witnessing a lot of traction and adoption. Leveraging this technology is essential, and widely used.

Is artificial intelligence a threat to employment?

AI adoption will impact one's current or future employment, which is a natural question, having started understanding the basics of AI. We want to share some of the relevant points and areas where one can focus to stay relevant, and this? is not an exhaustive list :?

• Automation of Routine Tasks: Move up the value chain for higher-end data and business analytics roles; do not continue to be a monthly MIS report preparer since? AI / RPA automation will sweep away this job.
• Enhanced Efficiency and Productivity: focus on AI model development, analytics, and process optimization skills. Learning about decision science will make you relevant since you have newer techniques.
• Customer Service Transformation: Mechanical jobs will vanish, but where human empathy, problem-solving, handling complex queries, and relationship building are involved, job will exist with an enhanced personal touch ( but learn to improve on soft skills).
• AI-driven Risk Management and Compliance: the ability to manage the tools and outputs, design enhancements, and interpret for decision-making essential. Learning the data and decision science tools would position you better. Also, stay put with multiple languages or tools; diverse knowledge will also be helpful.

What kind of skills will be in demand?

• Data Science and Analytics: Roles: Data scientists, quantitative analysts, and business analysts specializing in financial data.
• AI and Machine Learning: Roles: AI engineers, machine learning specialists, and AI developers focusing on financial applications.?
• Cybersecurity and Data Privacy: Roles: cybersecurity analysts, data privacy officers, and IT security specialists specializing in financial systems.
• Regulatory Knowledge and Compliance : Roles: Compliance officers, regulatory analysts, and legal experts with expertise in AI ethics and governance.
• Soft Skills: Roles: customer relationship managers, strategic planners, and roles requiring human oversight and decision-making.
• Technical Skills: Essential in the roles
• Understanding of Statistics and Mathematics
• Doing a course on Applied Data Science in Finance would help.
• Familiarity with data visualization tools like Power BI, Tableau, Qlik, Alteryx, SAS, etc.?
• Stay abreast of the latest use cases and adoption in the industry.
• Excel, Python, R, and SQL for data manipulation and analysis
• Understanding building scenario models
• Critical Thinking and Problem-Solving

To summarize, we have a basic understanding of AI and its application in finance, early adoption, and finance processes. We also got a taste of what kind of skills will be in demand and how a finance person should prepare for the next wave when challenged by AI. The scale will increase, roles will change, and jobs will always be there for the people who have adapted to the changes and transformed themselves. People who resist the change will be left behind, so please adjust and get the first advantage. As we advance, we will research and bring out more use cases in various industries focusing on finance and business processes, which would be valuable.?

Please let us know your opinion on our Newsletter.

Authored by : Subramanian Gopalakrishnan

Published by : CFO Bridge