For the second year running, SAP is using its TechEd global developer conference to showcase its support for citizen developers in the enterprise.
SAP is pushing its low-code/no-code approach as a way to help CIOs address the ongoing skills shortage by removing a bottleneck in the development process: a lack of developers.
The problem for many enterprises, says SAP CTO Juergen Mueller, “is just the sheer mass of requirements coming into IT departments.” When it comes to updating code or modifying reports from analytics tools, “Minor adjustments would not be hard to do, but because the IT departments are so booked, for multiple years often, stuff doesn’t get done, and then the business gets frustrated,” he says.
By enabling workers in other departments to develop their own projects, SAP wants to free up developers in the IT department to focus on the tougher stuff.
SAP’s commitment to low code
To that end, at TechEd this year SAP announced a “unified low-code/no-code development experience” on Business Technology Platform, its umbrella term for its collection of analytics, data management, AI, and integration tools built with harmonised interfaces and a common data foundation.
The no-code part of that unified experience is AppGyver Composer, developed by a small Finnish company SAP acquired in February 2021. Prior to the acquisition AppGyver had already caught the attention of DHL and Fingrid, the Finnish electricity grid operator, which use it to automate and streamline business processes.
SAP has now integrated AppGyver with some of its applications, although, says Mueller, “in principle you can connect it to any system,” either through the REST APIs and Odata (Open Data Protocol) interfaces in SAP Business Technology Platform, or through specific connectors.
The company offers a handful of such connectors already, but IT departments can also build their own. Even connector creation could eventually be democratised, says Mueller: “We are exploring how we might actually develop AppGyver with AppGyver.”
The low-code element of the platform is provided by an existing product, SAP Business Application Studio, which has been reworked to make it simpler to use. “We evolved it more into a low-code tooling to make it easier to access and take some of the hurdles away, to have an easier introduction into that world,” says Mueller.
There’s also a new “pro-code” tool to help developers experienced in SAP’s legacy programming language, ABAP, to write extensions to its modern cloud-based ERP system, S/4HANA, while keeping the core of the application clean to simplify software upgrades.
SAP S/4HANA Cloud ABAP Environment is based on an existing ABAP development tool for the Business Technology Platform nicknamed Steampunk. “It closes the last extension gap in S/4HANA Cloud to ease the transition from on-premises to the cloud,” says Mueller.
Daniel Newman, principal analyst at Futurum Research, says that the move into no-code and the commitment to development on Business Technology Platform are signs that SAP understands it needs to become easier to do business with.
“SAP recognises the need to play a more central part in the growth of the citizen developer and provide a more robust set of tools for customers to develop applications that can benefit from the data that resides on SAP and beyond,” Newman says.
AI: Trust, but verify
SAP is also aiming to simplify another programming task: that of creating conversational AI bots. Among the coming improvements showcased at TechEd 2021 are the ability to move bots between tenants — for example from community to production when ready to go live — and to initialise bots with memory and in a specific language.
Artificial intelligence (AI) is playing an ever-larger role in many enterprise applications, not just in bots.
SAP’s approach in workflow management is to put AI in an advisory role, supporting staff rather than substituting for them. It’s introducing a new recommendation service that learns from historical data or completed workflows. Configuration and training requires no coding, and can be applied to many different processes.
For instance, says Mueller, AI might be used to help in a vacation approval process to figure out whether multiple simultaneous absences in a finance department will result in one person being delegated the authority to make, approve, and pay a purchase order, and recommend whether a vacation request can safely be approved.
“We give the decision-maker insights and a recommendation whether that should happen or not. It’s not a one-and-zero, do-or-not-do recommendation. It’s like a percentage, how confident the system is that the approver should approve or deny, and then also it gives some reasons,” he says.
The implementation of AI across the enterprise is still in its early stages, Futurum’s Newman says. “Companies are navigating the utilisation of automation and augmentation to improve decision-making and streamline business processes.
SAP showed a strong empathic approach, which I believe will be highly adopted as we seek to maximise business performance by partnering human and machine to achieve the best possible outcomes.”