Practical Use Case for AI to Make your Job Better

One of the challenges with AI is that everyone wants to use it, but given it’s new technology, most don’t know how beyond treating it as an over-glorified search function.  I’d like to share an example of how I used AI so that it may inspire you to consider how you could adopt it – an old boss of mine who used to love saying, “let’s work smarter, not harder” and proper adoption of AI tools will help us all work smarter (not to mention will give us competitive edge!).

I previously worked in a very small organization with antiquated systems and limited reporting capabilities where the liquidity risks had never been appropriately quantified (beyond some basic regulatory metrics that is).  Through AI – co-pilot, specifically – I was able to create my own PowerBI dashboard which allowed me to compute structural interest rate risk, repricing profiles, currency mismatches, funding mismatches, concentration risks, etc..  The best part – because it was all automated – it could be updated monthly with limited human intervention, would take only a matter of minutes, and could be run by anyone on the team thereby reducing operational risks as well.  All I had to do was write some well-structured prompts, e.g.:

I have a data tape that has the following columns of information, I want to be able to load the data each month into a PowerBI database so that I can summarize the information and see trends in certain metrics over time (whilst allowing drilldown capability so I can see the underlying transactional data making up certain positions).  The file naming convention is always consistent each month where the updated file is named ‘MMM’ with each new month.  As a first step, how do I first create an automatic load so that I can have the new transactional data appended to the underlying database each month? 

This is a pretty niche example but think how powerful it is for someone typically used to relying on big-IT projects to get new reports built being able to do so themselves?  Not to mention that the process went much faster as I didn’t need to educate someone on definitions and metrics in order for them to build it.  (You may argue that’s not the best use of my time, but if you know anything about the summer heat in Dubai, you’ll understand why I didn’t mind spending a few weekends doing it!)  Having a tool that allowed us to explore in ways we never thought possible, meant we were asking questions we hadn’t even thought of before (and were able to mitigate potential financial risks with much more confidence and certainty). 

With AI, technology is no longer limited to the IT department.  However, if you do look to use any of the AI tools yourself, just make sure you do not include proprietary information or customer data unless you have the enterprise version of the tool (read: the greenlight from your IT and Risk departments). 

Thoughts and opinions are purely my own, no ChatGPT here.  The above is just a brief conceptual summary, for more specific details, feel free to use ChatGPT.  I’d suggest a prompt along the lines of:

I want to explore ways in which AI can help me and my team do our jobs better.  I am mainly responsible for [XYZ] and we employ the following systems to help us do so: [ABC].  Are you able to identify any areas where you can help us do the work faster, more accurately, or with less manual intervention?  Please sort the list by those which you think are the easiest to accomplish to the hardest and ask if you have any questions to help in your recommendation.