An Intellyx BrainBlog by Jason English for SAP Build
Let’s say you have some development skills. You can code a business application that calls APIs and talks to the SAP backend systems. You need pro-code tools to get your job done.
Or, let’s say you are not one of those developers, but you have business skills. You know what customers need, and you understand the business process that gets kicked off with each web transaction. You need low-code tools to get the job done with automation, while abstracting away both paperwork and technical complexity.
If you are in one of the abovementioned groups, then stay with me here. If you want to stay ahead of the market and build better applications, faster, maybe you should look to mid-code tools to advance your career.
Pro-Code, Low-Code, or the middle road?
Zen masters often talked about a middle way. Or, maybe it was a samurai thing.
Either way, we’re talking about charting a middle road through the software development process, somewhere between coding apps, connecting APIs and writing queries from scratch, versus using drag-and-drop apps to connect business process flows with data.
My colleague Jason Bloomberg penned an useful paper for SAP practitioners titled “The Future of Software Development” (free download here) which highlights how pro-code and low-code tools are starting to incorporate facets of each other:
“Many pro-code tools incorporate low-code capabilities, such as drag-and-drop API interactions, ‘boxes and lines’ for workflow design and data modeling, and wizard-based data integration.
At the same time, low-code tools have added more sophisticated features. Low-code vendors have always understood that hand-coding plays a significant role in any organization’s use of their tools.
Despite this convergence, pro-code and low-code tools serve different goals.”
While the goals of these toolsets may be divergent, what if both pro-code and low-code developers could collaborate in a ‘mid-code’ fusion environment where the two spaces converge?
The business perks of a fusion development shop
Line-of-business employees who work outside of development are often described as ‘citizen developers’—individuals who understand business context and could contribute useful application functionality to the business—if only they had the right low-code tools.
The SAP Build suite already offered visual drag-and-drop capabilities for low-code users to stitch together pre-built workflows and data from SAP backends and other integrated services.
With their latest release, citizen developers can ‘uplevel’ their business expertise in a mid-code way, with a genAI model trained on specific business verticals, and horizontal functions such as finance, HR, or supply chain, to allow natural conversational queries for guidance.
For instance, take a regional market development owner for an automotive brand. This business-user could ask and answer questions to design a mobile app for dealers to order new cars and replacement parts that talks to SAP inventory systems, and Ariba for purchase orders without writing a line of Java code.
But let’s say that app is about 95% complete—and still needs a capability to credit an employee identified in the HR system with a sale. At this point, a technical developer can jump into a shared project, with a few lines of ABAP or a change to an API definition, they can save the new adapter for use in similar processes.
Each new feature developers code and introduce modularly improves the relevance and utility of the low-code app design and process automation environment, for an innovation flywheel effect
Staying ahead of commodity work with genAI
There’s a lot of talk about genAI coding assistants and agents replacing developers to save money, but upon scratching the surface of any serious enterprise, that hype bubble pops quickly.
Enterprises never have a deep enough team of skilled business developers to deal with the modernization backlog they already have, so if anything, they are looking for genAI to help the developers they retain improve productivity.
“Including genAI in the developer’s toolbox will definitely save time, but perhaps not as much as people hope,” says Bloomberg in the whitepaper. “It will shift developers’ activities away from easily automatable, routine tasks to more sophisticated activities that organizations will want to leave in the hands of human professionals.”
Developers want to spend most of their productive work hours ‘in the zone’ of developing innovative solutions, so the most practical place for genAI to assist them first is as a co-pilot within their IDE, providing lookahead suggestions and potential error conditions while they are actively coding.
Determining the best approach to each unique problem is still up to the developer, but everything non-essential to the problem at hand could be optimized with genAI backed by a library of solutions and pre-tested components, from setting up infrastructure and permissions according to company policy, to coding responsive UIs for different devices.
AI has one superpower compared to human developers: it never gets bored, so it doesn’t mind tedious work. Developers can call on genAI to shortcut tedious tasks such as rebuilding existing integrations and writing documentation, while remaining fully in control of their own contributions.
The Intellyx Take
The future is going to demand a lot more out of us all. If we want to stay relevant, we don’t need to become superhuman technologists, we simply need to find better ways to work together and understand each other’s intent.
GenAI can augment human productivity with its understanding of the languages of both humans and systems, but it will never replace human ingenuity. The latest SAP Build which offers both pro-code and low-code teams a new ‘mid-code’ application delivery modality with AI in a supporting role.
When both sides of the technical divide can get involved in building solutions in a fusion development environment, hand-in-hand with automation and genAI, perhaps we’ll see that divide closed for good.
©2024 Intellyx B.V. Intellyx is editorially responsible for this document. At the time of writing, SAP is an Intellyx customer. None of the other organizations mentioned here are Intellyx customers. No AI bots were used to write this content. Image sources: Adobe Image Express (JE modified).
