Q&A Mark Lay – Nexus

Tell us about your partnership with Demarq. How did it come about, and what is it like working with the Demarq team?  

I was first introduced to Demarq in 2020 when I had recently started Nexus. We spoke about their work with UK Government, building bespoke applications, and particularly what they were doing with Optical Character Recognition. At the time, I was focused on building Nexus, but we stayed in touch as we both felt we would be working together at some point.   

I have been impressed by Demarq’s continued focus on practitioner-led development and the expansion of their app development capabilities using the Oscar framework. I’m certain Oscar can deliver significant value to our customers, and Demarq’s approach is perfect for so many of the organisations I engage with.   

My partnership with Demarq is primarily as an SME, bringing my extensive experience of working with Police and partners over the past 3 decades. With Nexus, my role has expanded into providing advice to both private and public sector customers. I have been working with Demarq to identify data challenges and new and exciting use cases for Oscar and will be helping them increase exposure within the sectors I now engage with. 

What was the problem you needed to solve, and what solution did Demarq provide?   

Whilst Oscar was initially built to extract financial data from bank statements, the Oscar framework allows us to solve a massive range of customer issues. Within the group of customers, I speak to every day, it’s clear there are many potential uses for Oscar.  

There are a lot of very clever solutions out there than can do amazing things: analysing data, investigating trends, identifying fraud etc. and there is a really strong focus on AI, machine learning and expensive, complex software solutions that can transform organisations and enhance their capabilities. But the thing that many of these solutions have in common is: they need the data to process in the first place.   

Data sits in many areas of an organisation: physical documents, claim forms, evidence forms, witness statements, invoices, receipts, emails, and dozens of other examples. Generally, this data is not easily accessible, or it requires manual input to transfer it to where it needs to be. This is a massive pain for organisations but often it’s not an issue that they are focused on resolving.   

Oscar’s capabilities are continuously growing! We have created new modules – including Invoicing, and Search and Locate – and are currently working on claim form extraction.   

We’re working with an insurance company that have a skilled group of data scientists. They’re running some clever algorithms but are missing valuable data enrichment as they can’t extract it from the claim forms and emails that it’s within.  

Another insurance customer currently spends hundreds of hours a year extracting data from invoices as part of the claims process. By automating this process, we can reduce time to decision, increase accuracy, and refocus valuable resources on investigating claims as opposed to carrying out data entry.  

What issues did you encounter when handling a large number of documents?   

The issues are pretty much the same regardless of the customer, sector, or use case. When you have many documents to process or lots of data to extract, your staff are focused on activities that are not beneficial to your organisation.   

Essentially, this is an Achilles heel in any downstream technology investment that requires access to that data. This bottleneck has the potential to significantly impact productivity, reduce your ability to be competitive, and increase your risk of data loss or integrity. Ultimately, this puts your business at risk, reputationally or financially.  

Data entry is a problem that no one wants to deal with. It’s not a high enough priority, yet it is the starting point for most downstream processes in the customer’s organisation. Being able to automate this process allows organisations to refocus their resources on key activities that directly affect their bottom line.  

How does Oscar help save money and resources?   

Depending on the document, the time taken to manually extract data from it varies, but we can estimate around 5 mins of data entry time per page. An organisation processing 30,000 pages a year would take around 1.5 FTEs to process those documents within the year, which would cost anything from £45k upwards. Using Oscar would see a cost saving of £40k per year – as well as a huge increase in performance, and a risk reduction. A company processing 300,000 pages per year could save over £400k each year. These numbers are indicative and quite conservative.  

How do you help customers implement Oscar into their current workflow?   

The driving concept behind Oscar is that it should be easy to use and deliver exactly what the customer needs. A customer can go from seeing a LinkedIn post about Oscar to importing documents in just a couple of minutes. It really is that quick and simple! For larger customers, who need bespoke pricing, the process is still extremely quick and efficient. Once the licence has been purchased, the customer can begin uploading documents immediately and, within seconds, have their data ready for investigation.   

What has been the biggest benefit of Oscar, and would you recommend it as a solution?   

The speed and accuracy of how Oscar can extract data is astonishing! Demarq has given a lot of thought as to how their customers will use the product and are passionate about delivering results. Conversations with larger organisations confirm that hundreds of hours can be saved, and hundreds of thousands can be saved in costs.