Introducing the OpenWrks Categorisation Engine: October Product Update

By Megan Harrison | Nov 16, 2018 |

We’re excited to announce not one, not two, but three huge updates to OpenWrks.  

Let’s get stuck in… 

Beta: Categorisation Engine

It’s been 12 months in the making. Conducted through numerous feedback loops with live customers. Analysing millions of rows of transactions to identify keywords and metadata...

But we’re happy to introduce the OpenWrks Beta categorisation engine.

Financial transactional data can be messy and difficult to deal with and the businesses that use this data need increasingly richer insight into their customer’s financial information so that they can make faster, more informed decisions.

Whether it’s quicker, more accurate income and expenditure evaluations or more sustainable, personalised and flexible repayment strategies, we’ve committed to providing our customers with the best category level insight.

So much so, that our analytics team have spent countless hours working closely with our customers during the closed beta phase to make sure we give them the insight they need. Every time.

As with all of our product, design and development methodologies, our data science team have deployed deep learning analytics. This involves supervised machine learning through continual live data feedback loops allowing us to rapidly iterate our models resulting in continuous refinement to give our customers the most accurately categorised financial transactions.

The OpenWrks categorisation engine groups related transactions and identifies keywords using a syntax language AI algorithm. Additional facts are established about each transaction group from the financial metadata. Our probabilistic model combines these facts and keywords to identify the most likely match, resulting in a categorised transaction.

So, we’re pleased to announce that OpenWrks customers can now get access to this level of categorisation for themselves through our OpenWrks API or the Surface Dashboard. 

Update: Surface Dashboard

Thanks to the feedback from our closed beta customers we’ve been working on making our categorisation easier to access and read for both users and developers alike. We’ve updated the transactional view in our Surface Dashboard to display categories so that agents in our customers businesses can see categorised financial transactions at a glance.


Plus, the Export to CSV option has also been updated to include the transaction category. Giving our customers a way to view, filter, and analyse their customers' categorised financial transactions.

We’re committed to providing our customers with the raw data and insight that they need to fuel their business decisions where ever they need it. So along with the Surface Dashboard, we’ve also updated our OpenWrks API Transaction endpoint to return the category for each transaction.


New and Improved: Developer Documentation

We’re a bunch of techies and we eat our own dog food, so we’ve been working hard to make the onboarding and integration of our OpenWrks API as seamless and easy as possible.  

We like making developers lives easier so the new and improved developer documentation has been designed to do just that. More detail, improved examples and easily digestible for all. Allowing developers to create secure, personalised, and performant products using the OpenWrks API.



What next?

We’ll continue to work on lots of new features. Is there something that would benefit your business or something that’d make your job easier, we’d love to hear your suggestions.


Want to find out more?

Test the OpenWrks platform at the Open Banking Expo on November 27th (25% discount tickets still available here.) Or get in touch to create your own OpenWrks lab environment and test and categorise your own data with us.



Join the discussion

comments powered by Disqus

The future of affordability assessments

Whitepaper: Our 4 predictions for changes to affordability assessments after the FCA updates its handbook.