On Public Cloud the waiting is still the cost: a Flexible Workflow that never started, an Application Job that runs overnight, an approval sitting in email until month end. Automate those and the cycle runs faster, the same way every time.
What is left over is the exception. We surface those sooner and put the context next to them, so the person deciding has what they need and the business can respond while it still matters.
A relaxed conversation first. If the answer is that you do not need us, we will say so.
Every step works. The process still takes a week.
Most of the elapsed time in a Public Cloud process is not spent doing the work. It is spent waiting: for Flexible Workflow to pick up the next approver, for an Application Job to run overnight, for somebody to notice that a purchase order is still in their inbox.
The same process also produces a different result depending on who ran it and how busy they were that day. Quality becomes a function of workload, which is why the errors cluster at exactly the moments you can least afford them. Month end is the usual example.
The exceptions that genuinely need a person are found late, usually by the customer or at period close. By the time anyone has the context to decide, the cheap options have gone.
of respondents name workflow automation and task routing as the AI use case delivering the most value, ahead of every other category.
SAPinsider, June 2026
The gain is in the handoffs, not the keystrokes. When a step finishes and the next one starts immediately, the cycle time falls by more than the sum of the tasks you automated.
A process that runs the same way on a quiet Tuesday and on the last day of the quarter is worth more than a fast one. Consistency is what makes the output trustworthy enough to act on.
Automation does not remove judgement, it concentrates it. The value is that your people see the thirty cases that need deciding rather than the three thousand that do not.
We do not implement Joule or SAP AI Foundation, and we do not build ML models. We also will not take on a process the business cannot describe: if nobody can say what the rule is, automating it just encodes the confusion.
More machinery is not better automation. We start at the bottom of this list and only move up when the rung below genuinely cannot do the job, because every step up costs more to build, more to run and more to govern.
Some processes should be simplified, merged or retired rather than made faster. Automating a step that should not exist just makes the wrong thing happen more often.
Standard SAP does more than most estates have switched on. If the requirement is met by configuration, that is the version your team can change without us.
Rules, workflow, events and scheduled jobs. The behaviour is defined, repeatable and testable, and it runs under the authorisations and audit trail you already have. This is where most of the work belongs.
Reserved for work that is genuinely non-deterministic, such as reconciling an exception queue where the next step depends on the case. It costs more per run and it needs its own oversight, so we use it where nothing below it will do, with a person still deciding.
Most SAP estates are running more than one of these at once. We work across all four, so where a process lands is decided by the rule above rather than by what we would rather build in.
The low-code option on BTP, for approvals and forms that sit across systems. It is quick to stand up, and it stops being quick the moment a process touches your own extensions.
In-stack and clean core, so automation is covered by the same authorisations and audit trail as the transaction it is automating.
Classic SAP Business Workflow you inherited from a conversion still runs real work. It is worth keeping until there is a reason to move, and the path is Flexible Workflow on Public Cloud.
For teams who would rather not have a vendor sitting between them and their own automation. Robot Framework and the Python ecosystem around it, running on infrastructure you control.
Automation usually has to reach something outside SAP. How we choose the platform, and the tools we work across, lives on the integrations page.
See how we integrate →AP automation with AWS, BTP, and a custom RAP application on ABAP Cloud.
Custom annual procurement planning to support compliance requirements in South Africa.
Replaced error-prone manual processes with custom barcode scanning and printing, so teams get accuracy and speed on the floor.
Longer form notes on the same problems, published on Medium.
Choosing the right automation trigger and using the Application Job Framework for Public Cloud extensions with BRF+ decision tables.
Read Blog ↗
Where agents genuinely help in ERP: the exception queue nobody enjoys working, handled with a person still deciding.
Read Blog ↗
Wrapping OData services in a single reusable iFlow, instead of building a new integration every time something needs to talk to SAP.
Read Blog ↗Most of what slows a business down does not need a model. It needs somebody to automate it properly.
Curious what a custom Public Cloud solution could look like for your company? →A relaxed conversation first. If the answer is that you do not need us, we will say so.