Defining real innovation in reconciliation technology

The world has changed significantly since we launched the Duco Cube service in April 2013: ‘best of breed’ hosted services have gained traction and are helping firms across the finance
July 27, 2016 - Editor

The world has changed significantly since we launched the Duco Cube service in April 2013: ‘best of breed’ hosted services have gained traction and are helping firms across the finance industry to cut costs, cope with regulation and simplify business processes.

Our story is about bringing fundamental research from the university lab directly to operations in industry and I would like to tell you more about how our innovations have brought Duco Cube to life.


Starting with a fresh perspective

When Christian and I decided to build a data reconciliation platform, our decision was shaped by the very poor experience of trying to apply existing solutions to a wide variety of different kinds of data.

The existing tools were built to solve very specific problems, for example reconcile bank balances, match FX trades, perform depot reconciliations, etc. They can be a good for dealing with one kind of data but they are poor at solving the wider problem of establishing data controls across an entire organisation. These tools also require a lot of effort to deploy and maintain and are unapproachable by non-technical users.

Today, it is clear that the rise of consumer technology is raising expectations that all members of staff should have access to powerful and usable technology. When you deliver this technology via a self-service model, you also empower your employees to help themselves, which is highly cost-effective and increases efficiency.

Our goal was simple: put the power to establish these data controls in the hands of business users who need them most and enable the business to cope with a wide array of different types of data, all on one platform.

It turns out that bringing consumer-like levels of simplicity and sophistication to the enterprise is incredibly challenging however. The slick consumer apps we have become accustomed to focus on simple business models. By contrast, the existing infrastructure, complexity of products and regulatory pressures in the finance industry conspire against this kind of simplicity.

Nevertheless, we have accomplished our goal of allowing non-technical users to reconcile any data by scrutinising every challenge facing them and innovating in order to keep things simple. Let’s take a look at some of these challenges next.


Comparing apples and oranges

The data you wish to reconcile is rarely in the same format. The traditional response to this is to employ complex IT tools to transform it but these are expensive and unusable by most business users. Our response was to create the Natural Rule Language, which traces its roots back to Christian’s PhD research.

This language enables business users to write rules that manipulate data, using English sentences. This makes it easy to overcome differences in data format, removing the need to use IT tools entirely. Since they are effectively English, the rules are clear to everyone and they are easy to maintain.


Getting results quickly

It often takes many months to see the first results when you use traditional reconciliation tools. To control implementation risk, companies resort to processes demanding lengthy business requirements documents and long testing periods. Business users are often only involved right at the beginning and the end of such projects and things often go wrong in the middle.

Our response was to invent a novel, self-optimising algorithm capable of matching very large sets of random data quickly and accurately, in memory. This empowers our users to take a much more iterative approach to reconciliation projects, as they get initial results immediately and can quickly fine-tune the matching rules in small increments.

This algorithm learns the form of the data and optimises itself as it works, allowing users to tackle very large data sets (e.g. millions of trades) and deliver meaningful insights in hours or days, rather than months.


Coping with complexity

Many systems dealing with complex data are themselves complex. There are too many buttons on every screen, making the user experience unpleasant and confusing. When was the last time you read the manual of a mobile app?

By contrast, Duco Cube is modern and simple. It takes advantage of the latest web technologies, is responsive and scales to mobile devices. We spend a lot of time streamlining the interface and ensuring it is clear and approachable.

If projects become too complex for our users, they are free to call us and we show up for free. We can afford to, because we spend significantly more money on usability upfront.


Coping with a changing landscape

The world is not a static place. Our customers need the ability to react to the market, so they need their partners to be able to deliver value quickly. Since Duco Cube is a service, we can upgrade all customers monthly, without them having to perform any testing themselves.

How do we manage this? Built-in regression testing that uses our customers’ production data to ensure results are identical before and after a software upgrade. If this is not the case, we immediately roll back the upgrade, so stable service is guaranteed.

The best part is that our customers don’t pay for new feature development on our platform, as the hosted model allows us to amortise the cost of development across all subscribers.


Closing thoughts

We are excited about the possibilities of the technology foundation we have built and the talented team we have assembled.

Duco has received a fantastic reception and growing recognition from the buy side, service providers and some of the largest banks in the world, often competing against larger and more established vendors.

Our philosophy continues to favour fundamental innovation over short-term solutions, which we’re convinced will pay dividends for us, our customers and the industry at large!

For more information go to

Most Viewed