In preparing for short-time work, companies usually do not lack software, but rather an overview of existing data, files, and responsibilities. This is exactly where a sensible entry point for AI lies: small, concrete, and without a large implementation project.
As economic uncertainty increases, topics that many companies would prefer not to address urgently come closer. Short-time work is one of them. As soon as it needs to be organisationally prepared or checked, it quickly becomes apparent how cumbersome such special processes can be in practice.
For the challenge is rarely just technical or legal. It is primarily operational. Employee data needs to be consolidated, timesheets checked, target and actual values compared, open points clarified, and various stakeholders coordinated. Especially in SMEs, much of this does not run in a seamless process, but is distributed across existing systems, spreadsheets, exports, emails, and manual additions.
The obvious reaction to this is often: Then we just need an additional tool for that. The second reaction is increasingly: Or just an AI agent.
From our perspective, both answers are insufficient.
Short-time work shows where the actual effort arises in companies.
Many companies already have software for HR, time tracking, payroll, or document management. Nevertheless, the organisational preparation for short-time work quickly becomes cumbersome. The reason is simple: the effort arises not only in the individual systems but especially in between.
Typical friction points are:
Employee master data is not complete or not up to date in one place
Monthly timesheets come in different formats
Planned working hours and reported hours must be manually compared
Entries, exits, or absences complicate comparability
Anomalies are only discovered late
Follow-up questions are handled via email, phone, or individual files
Management, HR, trustees, or payroll do not always work from the same data basis
Those who are familiar with such processes know: the problem is often not a lack of knowledge. The problem is a lack of structure in the interplay of the available information.
Why another software product is not simply necessary for this
Especially in administrative special processes, the reflex to look for an additional specialist tool is understandable. However, in many SMEs, this does not automatically lead to more clarity. It often only leads to another interface, another data state, and another process that needs to be maintained.
Short-time work is a good example of this. Most companies do not primarily need a new system for this. They need a way to bring together existing information more quickly, to identify inconsistencies earlier, and to address open points in a traceable manner.
The actual deficit is often not a lack of software, but a lack of connection between existing data, documents, and responsibilities.
And why SMEs do not need a complex AI agent for this first.
At the same time, AI is often discussed as if companies must start with agents, complete automation, and larger transformation projects. This is often the wrong approach for SMEs.
Because the reality looks different:
There is little time for elaborate planning.
Processes are not perfectly standardised.
Data sources are heterogeneous.
Implementation projects must justify themselves quickly.
The benefits must become visible early.
In short: SMEs usually do not need an AI architecture for short-time work, but rather pragmatic support for a specific process.
Not the big agent logic. But a simple, useful entry point.
Where AI can concretely help with short-time work.
The sensible use of AI does not lie in replacing legal assessments or autonomously deciding complex special cases. Its value lies in a much more practical area: dealing with distributed, inconsistent, and information-intensive queries.
Especially in the preparation of short-time work, AI can help, for example:
to classify different timesheets or file formats.
to sensibly assign employees or data records.
to recognise missing information early.
to better prepare target and actual values.
to make anomalies clearly visible.
Preparing queries in a structured manner for further examination
Compiling monthly figures in a comprehensible way for the review
The crucial point is:AI does not replace specialist systems or human review here. It complements existing processes where friction occurs today.
Thus, it does not become a spectacular agent, but rather an assistance layer that addresses practical bottlenecks.
Especially in short-time work, a grassroots approach is often more sensible for SMEs.
When companies are confronted with an issue like short-time work, they rarely need a major technological realignment first. They need a solution that makes a real difference with minimal planning and implementation effort.
This means in practice:
no months-long implementation project
no complete reorganisation of the system landscape
no complex agent logic as a prerequisite
no additional overhead for the team
A smaller, clearly defined entry point is often more sensible:
using existing data
better structuring monthly processes
enabling plausibility checks earlier
consolidating queries
making review steps comprehensible
Especially in SMEs, this pragmatic approach is often more valuable than any ambitious overall vision. Not because SMEs want less, but because they need solutions that are actually viable in everyday life.
What really helps companies with short-time work
From our perspective, the organisational preparation for short-time work mainly involves four things:
1. Overview
All relevant information should be collected centrally and transparently.
2. Comparability
Target and actual values must be clearly compared for each employee.
3. Early Clarification
Missing information, discrepancies, and anomalies should become visible early on.
4. Traceable Review
Open points, corrections, and approvals must be able to be processed in a structured manner.
If AI supports this effectively, it does not create another heavy system, but rather a practical relief in an already sensitive process.
Our stance at Mintsafe
We believe that short-time work clearly shows what many SMEs actually need when starting with AI: not maximum technical complexity, but concrete relief in a real administrative process.
Therefore, we do not see the value of AI primarily in grand promises from agents or in ever-new specialist tools. We see it where existing data, documents, and review steps can be made more usable with little effort.
Short-time work is a particularly illustrative example: a topic with real coordination effort, heterogeneous data, and a high need for overview — but at the same time an area where SMEs do not need new complexity, but pragmatic support.
Conclusion
When companies need to prepare for short-time work organisationally, it quickly becomes apparent how much effort arises in the gaps between existing systems. In exactly those areas, neither another specialist product nor necessarily a complex AI agent helps automatically.
Especially for SMEs, something else is often more valuable: a simple, practical AI assistant that better consolidates existing information, makes anomalies visible earlier, and supports the review process with minimal implementation effort.
Not another tool. Not primarily an agent. But a pragmatic entry point that helps in everyday life.