Overcoming Data Access Challenges and Detecting Fraudulence.

Discover insights from Mohammed Moafa, Data Governance Manager at neoleap, as he presents neoleap's success story with Incorta at Incorta No limits.

In his presentation, Mohammed Moafa elaborated on neoleap's effective strategy in overcoming data access challenges, resulting in a remarkable 59% success rate in preventing fraudulent activities using Incorta's advanced analytics.

Transcript:

We would like to present Mr. Mohammed Mahafa the data governance manager at Neolip, who will share in Corta's remarkable success story there, detailing their journey of overcoming data access challenges and detecting fraud.

Learn how Neolip leveraged innovative strategies and technologies to navigate complexities and drive meaningful outcomes in the dynamic landscape of data analytics. Mr. Mohammed, welcome the floor is yours.

Everyone.

My name is, Hammed Mahafa, debt, governance and strategy manager at New Lip, I have a diverse background, ranging from data science, to development, to consulting digital transformation, and I believe this, this enabled me to leverage technology and how to employ it to, as as a success, driver at Nuleep.

Nuleep is, a subsidiary of, Araje, the largest bank in, Saudi Arabia, and it's also one of the largest customers in Europe and Middle Eastern Africa for Encorta.

It's providing a a wide range of payment solution, the, our product, your pay is actually, one of the widely used products for Infintech that's used in Saudi Arabia.

And serving about four or more than four million users around the world.

New Lib is a fairly new, young company, and having about, more than four hundred and thirty employees, half of which or about two hundred sixty users actually utilize or use in Corta dashboards that are, developed at Newleaf using In Corta on a daily basis, those dashboards, as of now, we have about three hundred and thirty two dashboards, but the number is actually growing, giving the business need, and, we, we, we, I would say we can, we develop at least a dashboard every day, to to serve different business needs.

So, this dashboard actually get the data from more than fifteen data sources, of different types, and amounting to about four point two terabytes of data to serve all newly users across all, departments.

So some of the challenges we face, basically, or one of the major challenges we face is basically the use case specific, challenges, where we have to do some exploration trying to interpret whatever reports or whatever actually, happening in the actual words, say that you you're promoting a new product and you assume that it would be a success, but then you find out that the numbers or or the conversion rate is very low. So you need to explore this and this is a very, case specific, to to this. So this is one of the examples we have. Also, Some of the examples is basically the visibility and availability of, the data for all, users, for all operational department or simply for all, employees.

As for, reporting and, and, New Lip is a fintech company, regulated by Sam.

We have to submit reports to Sam on a on a regular basis And for this, it's, basically, it's it it was done in a manual way. We have some dedicated teams or usually at in any organization, they have some, teams to actually work on this on a daily basis just to generate this report and share it with Sam. This is one of the major challenges, faced at any event a company regulated by some, of course.

As for the marketing team, we need to ensure that we actually drive effective campaigns.

So where we actually engage our customers and deliver, what they actually need and want not what we actually want to sell. So, this by by delivering, these type of campaigns, is basically you ensure that you increase the customer, stickiness and loyalty, which is another challenge faced at any fintech company, giving the the that the currency we have multiple, fintech companies providing different offers and and, and and users could have could opt to either use your pay or any other, fintech company. So this is one of the major challenges, also, fraud, which is, I think one of the major topics, in in in the financial industry.

Fraud, detecting fraud or the the fraudulence behavior is basically a major challenge, for financial institutions.

And, usually, it's a it's it's a very, exhausting task to do with the with the reporting tools, the excel, you have to make some assumptions, And, basically, by the time you arrive at this report, fraudsters might have already committed tens of or hundreds of transactions. It would be too late by the time you arrive at that conclusion.

So, as for the consumers, We have a challenge that we given that we have more than fifteen, data source.

We needed to see where our consumers actually, are in terms of, in terms of, finance, in terms of demographics, in terms of any other, attributes or aspects that is stored in some database, that's not accessible by some other systems.

So, this is one of the challenges we have, and and and we we actually, saw the three sixty view of our customers.

As for the IT teams, or basically even the business teams, they're basically one of the challenges is how to have simplified visuals that would actually, drive to the point and, in in in a modern simplified way that's readable and understandable by any non technical user.

Last point is for, compliance which is basically, could be related to the fraud and the, summary reports, which needs to to have insights on a daily, weekly, monthly, and yearly basis.

And this is a challenge, especially if you are a young company with less than four hundred with less than five hundred employees.

So we need some sort of a miracle to to drive all of this and, submit your reports on time and be compliant with the PDPL and DMO, Sama, and all the other regulations.

So you need an army of, employees.

This is why we actually started using Encorta.

In Corta is not just a visualization tool.

We actually use it for operational and audit reports, as well as for the customer activity analysis and behavior analysis including fraud detection, and the integration of the other, platforms and data sources. To bring everything into one place.

Let's start with the first point for the operational and audit report.

Say given that, we already established that you have more than fifteen data source, wallet systems, CRM, a car transaction for audits, notification engines. So all of these data scattered across different dashboards and systems in different silos.

So we needed a centralized repository for all this data, a centralized location that would read all the data from all these data sources to be able to have a three sixty view, centralized, automated, and accessible by all team members from different departments and business lines.

Having in Corta enabled us to connect all these all these systems.

And given that, we we could automate, these reports especially for summer, for compliance purposes, also for, upper management. For example, you could get your report eight AM in the morning.

No need to actually have an entire team work on it to do the analysis and whatnot.

And this way, actually, increased speed and efficiency, of of our team and of our deliverables as well.

Some some of the things with that we were, that we're able to accomplish, over, using, in Corta is basically we were able to, find like twenty percent decrease in a card with no transactions.

Because once you spot a problem, you're able to find the resolution for that problem.

Encorta helped us actually reduce this.

And not only reducing the number, but all number of with no with no transactions but also increasing the issuance by sixteen percent, which is basically I'm I'm weeding out the, let let's say the cards that bring you no value, no revenue, with more active customers just because you had the visibility and you were able to actually detect that this behavior is happening and, you have some customers that are non financially active.

So once you once you, spotted this, of course, we at New Lip, we're able to reduce this by driving campaigns, by, actually having more having more targeted or addressing these points, So, by going deeply into the customer analysis and customer activity and behavior analysis, and subsequently, of course, fraud detection.

We had with the with the with the brilliant team, on at Nuleb, given that with their strategy and data architecture, AI machine learning, or, our develop developer teams we actually, built this structure where we integrated, in quarter integrate our systems within Corta on using, Apache Spark.

So we were able to actually develop our in house machine learning models that utilize all these computational power and resources to be able to, detect and report fraudulent accounts in real time.

So given that we have this architecture, we were able to complete cases, use cases in matter of days, not months.

Say that you have a business question.

I wonder what if?

You will know what if next week. Isn't have to wait until next year, next q, or, next month for the monthly report.

This actually gave us the edge in the market to win over our competitors.

New leap, we are a very, data driven organization.

And we believe, in data as an asset, and we treat it as so.

So we have a complete view, of what happening either within the organization itself or externally within the customers.

Cash in, cash out. We have you have the complete picture here. Once you have the complete picture here, Of course, you can make better decisions.

I'm not talking about like high level picture, but at the very, depth at the transaction level.

At the very low level, in real time, you'll be able to you'll be able to make a better decision and to detect any suspicious behavior.

So we conducted a use case last year, and, we say out of our millions of users, we detected some potential frauds that is using the machine learning algorithm we just talked about.

We discovered that we had thirty eight thousand, potential fraudsters detected.

Okay?

Out of out of those thirty eight k, There were actually fifty six percent confirmed and reported fraudsters.

And this was just using this algorithm to report the any suspicious, behavior.

So we actually went from reactively reporting fraud after it's committed to proactively detecting and reporting potential frauds to watch for.

So we prevent fraud even before it starts.

This is one of the use cases that or this is one of the things that could be accomplished or could not be accomplished. Had it not been, to the usage of of, this tool that and, of course, with the brilliant mind of our team that we're able to put these things together.

And to to to arrive at this very, innovative case.

As a result of this, newly now actually prevents fraud, we have, fifty nine percent of, of reported fraud, basically, this shows that New Lip is capable of detecting fraud before it happens and confirmed with numbers that fifty nine percent that you have, reported the actual fraud, which is basically a very high percentage, given that there is no, there is no like a clear cut of what fraudsters do or what defines a fraudster, other than behavior.

So, newly plow can monitor all customers' activities and, detect fraudulence because of this real time analysis and the transactional level, analysis as well.

As for the integration with other data management tools, of course, as you heard early in the presentation, that in Corta X now incorporates all of this as built in functionality. However, we at Newleaf were able to actually integrate with informatica, a stack.

Informatica actually have, the informatica data quality, for example, You have, enterprise data catalog where you actually monitor the metadata, and you monitor and check, the data quality and ensure that, whatever data you are reporting or you see in those dashboards, you can actually rely on. So with this integration, we were able to govern the the data aspect or the get the governance aspect, of, of this, integration along with the capabilities of Encorta. So as of now, as you see, we covered this process end to end.

Not only visualization and reporting, but also governance and quality checks.

So this is the high level architecture of the integration. We can see I'm not go going to the delve deep into the actual architecture, technically, but, you can see that you have this layer here for informatica where you actually monitor the data quality and, the metadata.

So imagine you have a user, you are a user, of, of this tool, and you need access to a dashboard, or you want to see a dashboard.

But you don't know the type or the what of the data in there.

Does it have this attribute you're looking for?

It might not.

So imagine you have a way that you can type a dashboard and see the insights that are included in this dashboard.

And you see the attributes that are covered in the dashboard.

Before even opening the dashboard and checking the dashboard.

And this would not be able without this integration with, EDC for, reading the metadata and populating the metadata into what we call a marketplace.

It's just like you're shopping for dashboards, literally.

So with this integration, actually newly is able to to, to say confidently that we make well informed decisions based on the high quality data we have And the insights that we get in real time, we act today, not tomorrow.

And without room to make decisions, We actually dwell in yesterday's glory.

Thank you so much.