Processes that run without you.
We automate the recurring steps that eat your capacity — with AI agents wired into the systems you already run.


Our clients
Insel Gruppe — Bern University HospitalHow an agent reaches production.
We start from the process and stop only once it runs under monitoring.
Measure the process
How often does the step run, how long does it take, what does it cost?
Connect
Access to the systems where the process actually happens.
Build approvals
What may the agent do alone, what needs a human confirmation?
Monitor
Error rate, cycle time and a kill switch from day one.
Engineers, not just consultants.
Technical delivery is led by Abhinav Gupta, previously an engineer at Emergent (YC S24). You talk directly to the people who build it.

Abhinav Gupta
Lead Software Engineer
Abhinav leads engineering at getSichtbar, from the platform and infrastructure to technical delivery. Previously, he helped rebuild Emergent's (YC S24) deployment engine with Temporal.
LinkedIn
Aatmik Panse
Senior Software Engineer
As senior software engineer, Aatmik builds our platform and AI agents for analysis and recurring work. At PhotoSage, he integrated private on-device ML photo search into an iOS app.
LinkedInAI Automation & Agents in practice.
Documented projects with figure, period and the work behind them.
+761%
more impressions in 3 months
How a Shopify relaunch led to more organic demand.
+315%
more conversions from AI assistants in 3 months
How a lighting store used Sphen to reach more buyers through ChatGPT, Gemini and Perplexity.
+115%
more impressions in 4 months
How a clear medical IT site structure doubled visibility in a regulated niche.
What you get and who it fits.
What you get
- Process map with time spent
- Agent running in production with system access
- Approval and escalation logic
- Monitoring view with error rate
- Operations documentation
A fit if
- The process recurs and can be described.
- We get access to the systems involved.
- Someone may decide what the agent does on its own.
Not a fit if
- You only want a report and no implementation.
- You expect a guaranteed mention in one exact AI answer.
- Your offer is not clear enough to recommend yet.
How working with us goes.
Four steps from the first call to ongoing delivery. The process is the same for every service.
First call
In 15 minutes we clarify your category, your goal and where you stand in Google and in AI answers today.
Analysis
Our platform runs real buying questions through ChatGPT, Gemini, Perplexity, Claude and Google and shows which providers get named and which evidence you are missing.
Plan
You receive a prioritised plan with sequence, effort and expected impact. Then you decide whether we work together.
Delivery
We implement, measure mentions, rankings and traffic every month and report in a shared Slack channel.
What our clients say.
Quotes from founders and managing directors we work with. Where a case exists, the number sits next to it.
“No fuckin way, Felix Ament is the GOAT. Within weeks, NICCOS was #1 on ChatGPT and Claude. Our organic inbound leads have exploded thanks to getSichtbar.”
+761%
more impressions in 3 months
“In the first month, the first request came in with the line: 'ChatGPT recommended you to me.' That is exactly why we use getSichtbar.”
+980%
more organic conversions in 3 months
Common questions about AI Automation & Agents.
Which processes suit AI automation?
Steps that recur often, can be described clearly and cost people time today: quote preparation, data maintenance between systems, reporting, first-pass triage of inbound requests. The more often a step runs and the more uniform it is, the faster the build pays for itself.
What happens when the agent makes a mistake?
Every run is logged and every action has an approval level. Critical steps only execute after a human confirms. There is a kill switch and a monitoring view with the error rate, so a problem becomes visible before it becomes expensive.
Does this need its own platform or licence?
No. We build into the systems you already have, through their interfaces. If an existing platform covers the process more cheaply than a custom build, we say so and do not build. The recommendation comes from the maths, not from our offer.
How long until the agent is in production?
Usually four to eight weeks for the first process, including measurement, integration, approval logic and monitoring. Most of that time goes into interfaces, permissions and edge cases rather than into the model itself.
Why start with only one process?
Because breadth is the most common reason projects die. Gartner cites unclear business value and cost for the 40% of agent projects cancelled by 2027. One measured process running in production gives you the number you need to decide on the second.
Does the knowledge stay with you?
No. We hand over with operations documentation and your team gets access to monitoring and configuration. You should be able to switch the agent off, change it and run it without us, even if we stop working together.
Talk to us about your visibility.
15 minutes, straight from the calendar. Before the call we check which providers AI names in your category and bring the result.

Felix Ament
You talk to me directly.
What happens in the call
- We show which providers ChatGPT, Gemini, Perplexity and Google name for buying questions in your category.
- We show which sources support those answers and where competitors stand.
- We name the evidence your brand is missing today.
- You get the next step with the biggest effect. No obligation.



