The grass IS greener!

close up photo of green grass with dew drops

Criminal activity is a reality of the world we live in, which means that combatting fraud, money laundering and terrorism financing is too. For any transactional company, this is paramount to reducing losses and minimising penalties such as higher fees for merchants and consumers. For South Africa, a significant consequence has been our greylisted status that impacts our economy and the ease of international business. The obvious answer is to implement robust systems to improve the detection, and therefore prevention, of illicit activities in your organisation.

With all the fraud prevention, anti-money laundering (AML) and combatting the financing of terrorism (CFT) solutions out there, we thought some clarity was in order to assist you in making the right choice for you. Of course, we’re biased but we’ve also done our research (no other company in Southern Africa offers graph-based fraud, AML and CFT solutions) and spent years developing the ultimate graph intelligence platform.

You can’t hide from graph

And there’s our blatant bias: graph technology. No surprise really, graph’s phenomenal ability to connect data, map networks and illuminate hidden relationships has transformed the analytics landscape. So, whilst other software suites use machine learning, AI and big data analysis, without graph they aren’t as effective at identifying complex relationships and patterns between data points. That means sophisticated offenders will still be able to find a work around because that’s what they do. To counteract that requires transparency, which is what graph technology can provide.

For example, cross-account collaborative fraud (where criminals work together across multiple accounts) is notoriously difficult to detect as it mimics normal trading activities. But the detailed analysis of a graph data model allows you to investigate the relationships between different entities and catch this suspicious activity before severe losses are incurred.

So yes, we rave about graph but that’s because this level of analysis is either not possible or computationally expensive using legacy database technology e.g. SQL (which still very much has its place). However, the reality is that it has technical constraints in processing and analysing large, connected data sets in real or near real-time. And, let’s face it, the criminals aren’t going to wait around for your systems to catch up.

Ineffective rules

Notes and coins on a table next to a candle and one handheld note burning.

So, there are glaring transactional volume and real-time analysis issues. But there’s also a problem with bog standard rules, which criminals know how to circumvent. Here are two real life examples:

Example 1

Recently, I heard a shocking story of how a friend of a friend was defrauded of R50,000 from her credit card in a 12 hour period. That’s money she doesn’t have and will have to pay back because she had given out her details for a transaction so her bank won’t cover the charges. Horrified, I asked how that was even possible. The fraudster debited her card in R5,000 increments, which wasn’t detected as anomalous behaviour. And there’s the real problem: a flawed fraud detection system that is ill-equipped to prevent flagrant card misuse (can you tell that I’m furious about this?). And sure, giving out your credit card details is a no-no but she was supposedly on a secure online site when she did so.

Example 2

We’ve had at least three instances in the last two years of under R100 fake monthly debit orders. Each time we are given the standard response that we need to contact the third party to stop it. Can you imagine the email if you ever manage to track down an address?

Email on a laptop:
Dear Mr / Ms Hacker
Please stop stealing money from my account. Thanking you in advance [smiley face]. Kind regards

On the last occasion, when asked, “How can it happen when I didn’t authorise it?”, the customer service agent responded that it’s a hack. So, the financial institution is aware that these things happen but do not necessarily have the ability to do anything about it. Again, the onus is on the customer to spot and then rectify the issue.

In this case, the systems failure is a rules and detection issue. A fake account drawing under R100 from multiple accounts is not flagged because the data is not connected in such a way that the financial institution’s system can pick up. Though our graph intelligence platform would because our graph specific algorithms detect anomalous patterns in transaction and account activity. This is the realm of “volcanoes” and “black holes”, which are paramount in AML / CFT and fraud prevention.

Diagram of a transaction network showing black hole and volcano configurations

We take it for granted that we can trust the financial platforms we use on a daily basis because we assume they have the robust detection and prevention systems in place to protect our transactions. But these real life examples show that’s not always the case. So, who can we trust?


Multicoloured light stream aimed downwards and splashing out at the bottom. Writing states: Unleash the power of graph! Locstat

I know I’m blowing our own trumpet here, but with alarming stories like those, I reckon we’ve earned the right.

We have spent years developing an ecosystem of next generation capabilities to fully unleash the power of graph. Locstat LightWeaver®, our graph intelligence platform, is a world’s first in using complex event processing (CEP) along with a strong computation engine to support a full complement of graph technology capabilities (graph analytics, graphML and graph visual analytics).

With each client, we configure Locstat LightWeaver® to their unique business to provide a highly contextual data model and graph intelligence system that relates to their specific needs. This model takes into account aspects such as customer needs, customer data, geography, as well as systems and operational environments. We map the data and engineer features according to these contextual nuances, which ensures greater accuracy and greatly reduces the number of false positives.

Locstat LightWeaver®

Empowering businesses with a world's first in graph intelligence

Our data source agnostic platform ingests data (structured, semi-structured and unstructured; batched or real-time) from all of your systems into a central repository for auto-graphing, processing and analysis. It is also API enabled and can integrate with your systems to operationalise outputs into existing workflows. A dynamic user interface provides dashboards and investigation tools, including data interaction and extraction.  Risk, opportunity and business insights can then be disseminated to various users as aggregated or automated reports.

This ability to bring everything together creates a more complete picture across your entire organisation. In terms of regulatory compliance, it also provides supplementary information to more accurately verify customer identity and behaviour – as well as to determine potentially fraudulent account holders or transactions.

Depending on your requirements, we onboard raw or curated data streams to improve the quality and pipeline these into the platform for event processing and rules action. The streams are enriched further with graph feature engineering and can then be re-pipelined in for deeper analysis.

As a light touch on your existing architecture, Locstat LightWeaver® is linearly scalable so, as your requirements and volumes grow, capacity is increased without disruption. But that’s not the only way we help you expand. We also don’t limit the number of rules and ML models that can be operationalised. Plus, we train your analysts to write, test and operationalise complex rules, so that you don’t have to rely on a third party to innovate and combat the everchanging fraud scenarios.

And if that isn’t enough reasons to pick us, let’s talk financial sense.

Better on your pocket

pink piggy bank surrounded by gold coins

Apart from the obvious benefits, with South Africa’s recent greylisting, strengthening AML and CFT systems is also an evident bonus. But, in reality, cost to company tends to win out over amazing, effective tech that delivers on its promises. So, let’s get down to brass tacks.

As a proudly South African company, our pricing is competitive compared to European or US owned solutions. To ensure excellent service, our subscription fee includes implementation and ongoing rule building until our clients are self-sufficient. It also covers maintenance and support so there are no additional consulting fees that diminish your ROI. Of course, it also helps that an implementation can deliver value within weeks.

And for my final trumpet blow.

We don’t just have an incredible platform that can solve the complex problems you’re dealing with, reduce losses and costs, minimise impact to customer, tighten up regulatory compliance, adapt to changing scenarios and scale as you grow. We open up possibilities by empowering your team with the skills to use it and deliver more in our fraud detection, AML /CFT solution at a lower cost than our closest competitors.

When you’re unique, there is no competition!

Sunrise over a field of grass. Face a new day in greener pastures. Locstat

You might be interested in …