Asset Control to Add Data Model Customization to Ops360

Data management software provider Asset Control is looking to add additional model configuration flexibility to its Ops360 data workflow platform, which was launched at the beginning of March.

Mark Hermeling, CTO at Asset Control, tells WatersTechnology that while the platform allows users to manage their data workflows for things like data validation, lineage, and propagation, users cannot currently customize the data models, so Asset Control has been doing so behind the scenes. With their next release, the vendor will provide tools for users to configure those models.

“It [currently needs] configuration behind the scenes. What we’re releasing next is a full suite of apps and tools which allow you to structure your models, your views, your business process models, [and] the rules you apply so it gives the user dynamic access to that data,” he says.

The web-based data operations platform was built using a microservices architecture, and gives users full control over the data acquisition and mastering process. It allows users to browse, process, and validate trade data more easily than traditional manual practices.

“With microservices, it’s very modular—it’s all stateless—so it is scalable and allows for [easier data] distribution,” he says. “If something in the solution goes down, all the other bits keep working.”

Asset Control has taken that premise a step further by incorporating what’s known as a micro frontend—an approach to web application development comprising small frontend apps. Essentially, the microservice can be broken down into domain-specific micro frontends that are self-contained and developed, as well as deployed independently. 

“So what you see in the UI [user interface], are individual components that plug into microservices, but they’re represented in micro frontends. So if one of those services goes down, only part of your UE [user experience] would be affected. That means that from a user’s perspective, there’s a lot of resilience,” he says.

Asset Control is now also starting to look at generating statistics around trade data quality, looking at whether the data is correct, whether it arrived at the right time, and how long that entire process took. 

“In the old days, you take a group of data, you slice it, and you get your statistics, and you can combine that information to draw a conclusion. We’re doing that based on the data model, and we’re saying: For US equity data, how did the data flow through our system, and on a statistical basis, how often did it need to be corrected? So we’re deriving those statistics, we’re going to present it in our tool and make it accessible to the user,” he says.

Beyond that, Asset Control is looking to apply machine learning techniques that will propose changes and ‘what-if’ scenarios to data processing models. 

“So if you have your data lineage in your model … [you can see] how the data flows through your system, [and] you can heat map your data processing model. Then, you can actually use machine learning to take those models, see what works, make adjustments and actually propose changes. And instead of proposing to change and just saying ‘change this,’ you can actually produce the same statistical report with ‘what-if’ scenarios. I think that’s our longer-term goal,” Hermeling says.