“If you want to drive digital transformation, if you want to overhaul the way your business works, you have to start with its data. Being able to operate on that data when it can’t even talk to one another is the very first challenge that most enterprises encounter.”
These are the observations of Adam Glaser, Senior Vice President of Product Management at Appian. Dynamic Business sat down with Glaser to discuss how, at a time when Australian organisations are drowning in data, their biggest challenge is not the volume of their business information but its fragmentation.
Data silos restrict the usefulness of organisational information
Operating a business using isolated legacy systems restricts access to and integration of data, a significant business challenge, according to Glaser. Siloed data results in inefficiencies, redundant efforts and missed opportunities. Further, being unable to consolidate and analyse data from various sources in real-time hampers decision-making processes and impacts business performance.
“Siloed data has an impact on every business,” said Mr Glaser. “Typically, what you have are sovereign data systems that were built for a singular purpose. That’s all well and good if all you’re doing with that data is CRM or accounts payable or customer tracking. However, if you want, for example, to understand the entire life cycle of your customer, well, now that business process spans all those different systems.
“So, now you’ve got the different data types, different ports and protocols, and different security rules. You’ve got owners of those systems who typically are very territorial. They’re protective of their systems, a little crotchety maybe, when a new application developer comes along and says, ‘Hey, I’d like to access your data as part of this application I’m building to help transform this other part of the business.’ Systems owners are going to say, ‘No, no, no, no. All you’re going to do is mess it up.’ We encounter that type of resistance all the time.”
DALL·E 3
How data fabric differs from other data management solutions
There are several technologies available to help businesses overcome the access challenges presented by data silos.
“Data lakes and data warehouses tend to aggregate data into a novel location, which does away with some of the challenges but introduces new ones because the data is now historical; it’s not real-time,” said Glaser. So, it’s good for analytical purposes but not for transactional purposes. The way they’re built is that you still run your business through siloed systems, but then, on some schedule, some batch job, you pull it together for the benefit of more cohesive reporting, which is fine if that’s what you’re interested in.
“But if you are looking to turbo-charge the overhaul of business processes, you can’t use a latent and out-of-sync data source, a historical data source, just for reporting purposes.
“So, then you get to data mesh and data fabric. Both of them are architectures, but there’s a key difference.
“With data fabric, what we’re doing is creating a virtual overlay, basically a virtual agility layer on top of existing systems. With this approach data stays where it is, and the platform is tied into those data sources so that they can be addressed inside the development platform as if they are one cohesive data object.”
Glaser reported that over 94 per cent of new applications built on Appian use data fabric. “It has become the standard way to address enterprise data,” he continued. “The benefits are just so strong, and the drawbacks are so few, that the only reason we don’t see 100 per cent adoption is people usually don’t go back to legacy applications and update them. They’re looking forward to how they can use new technology in the future.”
Data fabric is not a magic bullet
Glaser was quick to acknowledge that a data fabric is part of a solution, but it’s not a magic bullet.
“If the data is inaccurate, a data fabric is not going to fix it automatically. It will provide a single pane of glass into that data for understanding and maybe make it easier to identify that the data was wrong in a way you couldn’t possibly see before, but there’s a shared responsibility.
“We rely on the data owners, and we rely on knowledge workers. Effectively, these applications are by them, for them, to help improve parts of their business. If they are putting data into the wrong system or connecting to the wrong data, a data fabric isn’t going to know that. It’s only going to facilitate better development experiences or better access to that data.”
Highly regulated industries reap the benefits of data fabric
Data fabric adoption rates are particularly strong in highly regulated industries, and Glaser noted that many of Appian’s Australian customers are in financial services, life sciences, insurance and the public sector.
One example is the Victorian Office of Public Prosecutions (OPP), which recently partnered with Appian to deploy a platform that leverages data fabric and process automation technology to accelerate prosecutions.
The OPP’s previous case management system presented multiple challenges, including high maintenance costs, security risks, limited integration capabilities, fragmented data and a lack of scalability for future growth.
The development of Amicus on the Appian Platform brought together previously siloed legacy systems to streamline operations, reduce administrative burdens and automate processes to enhance productivity and optimise service delivery.
“Data is so important for public prosecutions,” said Glaser. “We are talking about data on crimes, data on criminals, and data on victims.”
Amicus has increased the amount of casework done per unit of time and decreased the number of manual errors occurring within Victoria’s largest criminal practice. Amicus also provides real-time reporting to prosecutors to help them better serve justice. The OPP anticipates that Amicus will deliver a productivity gain of over 10 per cent annually.
DALL·E 3
How data fabric enables Artificial Intelligence
When asked about the role AI plays in the functioning of data fabric, Glaser flipped the question to comment on the role that data fabric plays in AI.
“AI is nothing without data. Data is the foundation of getting any enterprise business value out of AI,” he observed.
“Until you have a technology like data fabric that gives you the comfort and confidence that you can address your business data in the way that your business users think about it, that you’re actually able to secure it so that you know that a piece of data that you’re sending into a large language model or to an AI algorithm is already scoped to be secure for the user, you would be hesitant to bring AI into the enterprise.
“We’ve seen that data fabric is an enabling technology for AI in the enterprise because now that you’ve got this solid foundation of data underpinning your applications, putting AI on top of those applications or within those applications, whether it’s to help with workflow—so structured/unstructured document classification and extraction, or summarisation, or knowledge management or generation, correspondence generation or things like that—until you had the data fabric there, you would really struggle to get the right data to the right AI model with any confidence that what you got back was trustworthy.”
The future of data management in Australia
“I see a lot of innovative customers in Australia,” Glaser observed. “I see a lot of customers that are trying to push the envelope, both on data and in big data use cases, and on AI and the promise of AI in terms of productivity either for knowledge workers or the promise of straight-through processing of some of these large-scale applications that don’t even touch knowledge workers until or unless some type of escalation needs to happen.
“And data is a huge part of that. I think we’re going to see increased demand for scale, increased demand for performance, increased demand for being able to address different types of data sources, not just relational, not just API driven; maybe streaming data sources, big data, cloud data sources, maybe even new types of data sources that we haven’t seen yet.
“Any architecture or any platform that wants to participate in workflows of the future is going to need to be able to address those data needs and those data sources, or they’re going to be left behind by newer technologies that can.”
Glaser said that what he likes about Appian’s Australian customers is they’re very solution-agnostic.
“They are outcome-driven. The incumbency of a particular data management approach, or the incumbency even of a vendor, matters less than the potential of a new approach or a new platform or vendor to help them achieve their goals.
“They’re very goal-oriented. We see a lot of our Australian customers with board-level OKRs, outcomes they’re trying to achieve, and SLAs that they’re trying to measure to, and they are relentlessly pursuing them with every possible technology that might help.
“And so, anyone that wants to be relevant in those innovative industries, those innovative customers, needs to be able to keep up with the demand.
Glaser also observed that Australian customers are not looking for vendors rather for technology partners.
“They understand that they’re pushing because of how goal-oriented they are and because of how technologically progressive they are. They’re sometimes pushing the bounds of what can be done in AI, or the Cloud, or data, and they’re looking for partners that can both help them manage their expectations but also help them chart a path to get them where they want to go. They’re not locked in on one particular approach, and they’re willing to try many different things very quickly.
“There’s a lot of innovation, there’s a lot of experimentation going on, even in industries like financial services, that you wouldn’t expect, at the technology level, because they have real-world challenges, they want to drive real-world efficiencies, and they see technology as a key part of it.
“Data fabric is an excellent part of what the total population of the solutions space is going to be. AI is quickly coming on board as another player in that space, and businesses are looking for more. They’re looking to see what else can help them achieve their goals.”
Keep up to date with our stories on LinkedIn, Twitter, Facebook and Instagram.
Dynamic Business sat down with Adam Glaser, Senior Vice President of Product Management at Appian to discuss how data fragmentation, not volume, is … Featured, Data Management Dynamic Business
