t cht lk
t cht lk
Everyday AI means everyone knows what to do in the case of , say , missing data .
already in full ascent , but transformation programmes are not without their challenges .
It all begins with data . Data on customers , data on markets , data on products , data on transactions . And as industry consolidation has stitched together disparate brands – Abu Dhabi Commercial Bank , Union National Bank , and Al Hilal Bank become ADCB Group ; National Bank of Abu Dhabi and First Gulf Bank become FAB ; NBD and Emirates Bank become Emirates NBD ; and so on .
Once we have established data access and governance , we must address skills , the shortage of which is a major obstacle to overcome in any analytics programme , even in the financial services sector , where data literacy levels tend to be higher than average . But not every employee is a data expert , and business-embedded analytics , what one might call Everyday AI , calls for an update in mindset , followed quickly by upskilling and change management .
Everyday AI means AI , every day . Every employee thinks about data , how it is collected , how it is stored , how it is accessed , and by whom and for what purpose . Everyday AI means everyone knows what to do in the case of , say , missing data . Do you use proxies or estimates or something else ?
The resultant companies have had to homogenise their data and smooth out their legacy processes to ensure those that need access to data can get it . But they have also had to contend with local privacy laws like the UAE ’ s Personal Data Protection Law of 2021 and the EU ’ s GDPR . And they have had to consider the strain on infrastructure and the absence of central warehousing for data .
Missing prices for traded instruments on a particular day may be suitable for estimates , which can be used to guestimate margin calls and risks , for example . But in other cases , estimates could have an adverse effect on decision making . Employee training , collaboration , between risk experts , domain experts , and data scientists , and governance will make the difference when such calls have to be made .
The best way forward has been to create secure staging areas for the testing of analytics models that are monitored through strong governance . Over time , data can be categorised by type and criticality , from which we can derive its priority in backup and recovery , and its level of sensitivity , from which we can derive the roles that can access it .
As the financial services industry ’ s AI journey moves beyond investment teams , first to the plate , because of their ongoing quest for market insights , banks carry with them important lessons regarding AI ’ s usefulness . Institutions have discovered that the AI toolbox can help with more than just predictions and new business models .
76 INTELLIGENTCIO MIDDLE EAST www . intelligentcio . com