We use AI where it gives your team hours back.
We introduce AI applications in your company and connect them to your existing systems. Every automation first proves itself in a trial run.
Our stance
AI impresses in every demo. We are interested in what remains of it after three months of everyday work. That is why we measure every automation by the time it saves.
Services
Where AI saves you time
We use AI in operations and in administration. Most of the examples come from our background in network operations.
-
01
Analysing operational data
The system learns your normal state from the metrics. Deviations therefore stand out early. We often spot them before the first complaint.
- Metrics
- Normal state
- Early warning
-
02
Automating processes
We automate processes with fixed rules. These include configuration checks, updates and reports. A test run comes before every deployment.
- Configuration
- Updates
- Reports
-
03
Supporting troubleshooting
When a fault occurs, AI sorts the alerts and shows the likely cause. You also see the data behind the suggestion.
- Alerts
- Cause
- Evidence
-
04
Applications in administration
We introduce AI applications for your administration and connect them to your systems. Typical tasks are pre-sorting incoming enquiries and preparing recurring replies.
- Enquiries
- Replies
- Interfaces
Process
How an AI project runs
- 01Discussion
We identify the tasks that cost your team the most time.
- 02Selection
We recommend the processes where the effort pays off. We advise against the rest.
- 03Trial run
The automation runs alongside and only makes suggestions. Your team still makes the decisions.
- 04Handover
The automation takes over once it reaches an agreed hit rate. You can switch it off at any time.
Step 03 in detail
Suggestions first, then decisions.
During the trial run the automation does not act on its own. It makes suggestions and your team decides as before. We measure the agreement every week.
You set the threshold for the handover. After that every action is logged and can be reversed.
| Week | Agreement |
|---|---|
| 1 | 58% |
| 2 | 71% |
| 3 | 79% |
| 4 | 86% |
| 5 | 90% |
| 6 | 93% |
| 7 | 96% |
| 8 | 97% |
From week 7 the agreement is above the threshold. Since then the automation has taken over the process.
Self-assessment
Will automation pay off for your process?
Think of a process that regularly costs your team time. Four questions show whether it is suitable.
Does the process occur at least once a week?
Does it follow fixed rules that can be written down?
Is the data you need already available in digital form?
Do errors in this process cost time or money?
In practice
From our projects
Our projects range from engineering firms to aviation.
-
Electrical engineering firm
- Result
- Processes run through faster. The team gains time for the actual engineering work.
- Task
- Advice and implementation for automating internal processes.
-
Aviation engineering company
- Result
- Leaner workflows and a higher level of IT security.
- Task
- Support with the use of AI and advice on IT security.
Questions
Frequently asked questions
Is your question missing? Call us or write to us.
Does our data leave the building?
Does the AI make decisions on its own?
Which models do you use?
What happens if the automation fails?
Other areas
Talk to us about your processes.
Typical reasons are a process that takes several hours each week or a fault whose cause takes a long time to find.