Appian World 2023: How low-code will evolve with AI 

What is a data fabric?

A data fabric is a virtual database that looks at the different data sources across an enterprise and sits above all those data sources to enable users to read or write to those sources in a way that unifies all data – it’s a means to discover, unify, secure and optimse a whole range of data objects that may exist in a variegated distributed data landscape.

The alternative to data fabric is what Calkins calls the black hole of data – this is a scenario where an organisation has to surrender a large amount (or all) of its data to a single data operator… but of course, that process means the organisation itself loses control and freedom in terms of what it can do with its information streams.

In easy times, it’s no problem to scale, you just write another cheque, noted Calkins – but that reality (in the post-pandemic world with infection, inflation and invasion all around us still) each workflow team has realised that it needs to do more with less. This reality is central to the way Appian is focused on productivity… and that productivity boost will come from many places, but a central engine will come from AI.

“Our AI philosophy stipulates that AI has to be easy to use in order to be valuable – AI is only as useful as the practical benefits that you can get from it,” said Calkins. “The value of AI (and this is of course a fundamental truth) comes from the data…. and our philosophy is that you [the customers, the users] should keep your own data and not give it to use in order to train AI for application and process-centric use cases.”

All of this serves as the backdrop to the new Appian AI Skill Designer technology which the company has launched this month. But when it comes to AI, we need to think about two core shape or approaches says the Appian CEO.

Public AI vs Private AI

When it comes to AI as a discipline, the Appian CEO sees it sit in two specifically different streams.

Calkins describes public AI as AI where an organisation has to offer up its data to an AI design specialist and/or use AI that has been trained on third-party data, which will logically never be as close to an AI model that was developed on a firm’s own datasets. In this format, an AI model in the public space can be used by a firm’s competitors. Conversely, private AI is a situation where every organisation can train its own AI internally so that it is the totality of the benefit of that AI model is retained within the organisation itself – Appian itself (if it were not obvious by now) is solely focused on private AI.

Thinking about how organisations should now use AI, Calkins views it in a specific way. He reminds that AI can help us make decisions, but offer options for humans to be able to decide what actions should be related to some (or all) of the decision advice being offered, we should think of AI alongside us humans. Because of this, the Appian CEO says that AI should be a great band member, but never a solo artist.

Perfecting processes

Referring to other product announcements tabled this week, Calkins pointed to Appian process management technologies that essentially works as low-code process mining. Offering customers the ability to detect inefficacies in the Appian processes they are running, teams can act upon the recommendations (which could for example see some tasks offloaded to an RPA bot) the technology might offer to make processes work better.

Appian VP of product strategy Malcolm Ross took over for the section of this keynote to cover what he calls out as four key areas of focus for Appian.

• Data fabric
• Automation
• Total experience 
• Process mining 

Appian’s Karina Buschsieweke hosted a session focused on Continuous Improvement (CI) by looking back at the history of Toyota using ‘kaizen’ also known as just-in-time manufacturing.

Annelise Dubrovsky, Appian VP product management also spoke more on Appian Data Fabric to add to Calkins previous exposition. Thinking about how process mining could help manage inbound communication from customers, Dubrovsky took the audience through Appian designer technologies that a company might use to define, construct and create a data model based on the information sources (in this case, email) and create a means of viewing and managing from that point forwards.

What low-code did next

A comparatively long single-block keynote session, we’ll close this reportage off here with a note to suggest that Appian is a company that has grown (it moved office locations on the Virginia-Maryland border back in 2018) and added to its workforce significantly. The company has also used its annual Appian World conferences to detail major platform-level developments and even moved forward to a point where it no longer makes too much fuss around the low-code tag that we have known the organisation for up to this point.

So then, just remember… it’s not low-code as a sole entity or technology proposition, it’s low-code as a substrate behind orchestrated end-to-end process automation driven by an increasing amount of AI-guided acceleration for software application development.

Appian CEO Matt Calkins: A vision to (positively) disrupt with a data platform product capable of changing markets.