Making Big Data Actionable Data

How Sphonic is bringing visualisation into the world of compliance to make the agent’s life easier.

By Andy Lee, – Founder Sphonic

Over the past 8 years at Sphonic we have focused primarily on real-time data processing. Providing access to big data and extracting, in real-time, what information is needed by our clients to enable straight through processing, to make automated decisions for KYC, AML and Fraud Prevention on their customer-base.

We have often talked about processing only the data required to fulfil the purpose of decision making and identifying the good first. Why? When searching for a needle in a haystack, sometimes it’s easier to reduce the size of the haystack by extracting big chunks of the “good” first up.

Now through product development we have also moved into the world of “back office” process reviews in the event anomalies are present at client portfolio level, which has required a 180-degree turn in terms of extracting what is left behind after the good is taken away.

When processing hundreds (if not thousands) of data points, or actors, a key challenge is how to display the information to a reviewing agent?

To put it another way, when you take your car into the garage, you give the mechanic an indicator or clue with respect to the problem, e.g. “there is a strange sound coming from the front near side”. Post inspection you want the mechanic to tell you what is wrong, not provide a full list of what is right with your car.

Upon receipt of the information you can then make an informed decision, which in turn will have consequences – namely the cost to remedy or searching for a second opinion (escalation in the world of reviews). The same principle applies to manual reviews in relation to suspicious applications or transactions.

In version 1 of our Case Management system we built a user interface that visualised all of the information processed in the Workflow Manager decision engine, with data categorised for ease of understanding. However, based on valuable client feedback, our customers are now looking for enhanced user navigation, so we have decided to double down on this in our next version.

In version 1, agents could prioritise not just on pre-defined strategies, but dynamically based on the risks and threat level at any point in time, which in itself created operational efficiencies. With version 2 we are doubling down on these efficiencies with faster and more accurate views of specific cases.

So what did we do? We already had the ability to segment data (tick) we already had the ability to create hierarchies within data (tick).

What was missing was the ability for an agent to choose what to display on a case by case basis – namely the relevant data for review from all of the actors available whilst dynamically building up an accurate and relevant picture in front of them.

As a result Sphonic has developed, and will bring to market in the next month, version 2 of our Case Manager product, where agents will be able to build their own “view” creating a bespoke story of the event taking place with a focus purely on the data of interest.

Various indicators (as with the mechanic example above) will be obvious and can be interrogated further using the new user interface.

We believe the power should lie with our clients rather than providing a limited out-the-box solution that can’t be tailored to the client’s specific preferences. Our clients determine what data is processed, what is available for display across all cases and what they wish to see displayed for any given case. Data should neither be forced on our clients nor unimportant / irrelevant data processed, which ultimately would be detrimental to both time and focus. This approach by design enables the data to fit into clients existing processes (and even enhance them) rather than the data forcing detrimental process changes client-side, due to systemic limitations.

For the majority of businesses Compliance processes and policies will already be in place and decisions driven by them. Sphonic focuses on the ability to meet existing (and new) requirements as simply as possible, serving up tailored information focused on enabling accurate decision making by agents, ensuring both speedy and accurate review outcomes. This approach ties into the new Case Management release where we have enabled summary level information to be complemented with focused data, and if further interrogation is needed, further deep-dives into all vendor data points are available through agent driven actions.

Sometimes “simple” is best. Data needs to be captured and processed, data needs to be identified for good and bad actors and data needs to be accessible, understandable and auditable. All of these key drivers along with valuable client feedback have helped us shape our revised offering. Besides V2 we will continue to listen, learn and enhance further versions through our continuous development programme.

Our system is available for multiple use cases from new customer on-boarding to suspicious transaction reviews all in line with AML, data protection and within industry regulations. The key is to enable the much-valued single customer view by joining up data both within and across multiple customer events, whilst enabling access to all key data points and displaying – via a user driven journey – what is of most importance.

Please reach out on LinkedIn or via if you would like to learn more or see a demo.