Products

What solutions do we offer

The Plausibility Engine is a unique proprietary machine learning algorithm that we have developed for the task of finding implausible data entries in structured data.

FNG Compare is a tool that lets you easily compare report results between different Front Arena environments.

What are the advantages of the Plausibility Engine

It learns from a past dataset that is assumed to be correct and then identifies anomalies in new data. It can also suggest what needs to be changed and if “the anomaly” turns out to be correct it can add it to its model to learn from this.

There is a variety of use cases where this tool can be applied. Data is the fuel for our digital world. Our analysis, our predictions and our decisions depend on the datasets we collect. Even with the smartest techniques data quality issues can lead to sub-optimal results.

Can artificial intelligence help us in refining the data

ai

Refining the data is a tedious task. Controlling for errors requires anticipating them. Humans can look at a case and conclude that it looks strange but they are overwhelmed by large datasets. This is why we asked, can Artificial Intelligence help us here?

The Plausibility Engine takes on this task with a minimum input from the user. It is a tool based on novel AI algorithms that looks for anomalies in datasets analyzing them from different angles.

These anomalies may be data quality issues. Maybe a sensor used for measurements is going wild or users are making data entry errors. Maybe someone is trying to commit fraud but haven’t quite thought it through.

Or maybe these are genuine anomalies, in which case you either want to take a closer look at them or you want to exclude them from your modelling.

Why choose the Plausibility Engine

With the Plausibility Engine you have a tool that can be used in a variety of settings. It is completely agnostic meaning that it only learns from the datasets it is provided with. It can be trained with clean data for out-of-sample validation or it can look for data anomalies within sample. It is easily extendable and scalable and can be incorporated within your processes and system architecture.

But it gets even better. The Plausibility Engine can also be integrated within your system architecture to do an on-demand data check. A use case developed to work with a trading platform for a banking customer can be seen here.

Make a consultation with us

We are excited to hear about your project and ready to discuss the best solutions for your needs.

Why choose the FNG Compare

It can be used during Front Arena upgrade projects to identify differences in calculated values that you need to focus on. The comparisons can be automated to redo after every refresh of the environments or new hotfix provided.
It can also be used to do regression testing when new packages and configurations are introduced.

This Can Be:

  • Code changes
  • Changes to the market data configuration

  • Changes to yield curves

For example business users can explore what valuation changes would occur if they introduce a new benchmark to a yield curve.

 

We wouldn’t claim that this is the only tool existing out there for the task but here is why we believe that our tool is better:

  • Standalone tool with zero changes to your Production code base

  • Intuitive and customizable GUI for your business users
  • Better performance and scalability through parallel processing

  • Ability to save your comparisons and run them as tasks in the background

Data analysis