What if it were faster to automate something forever than to do it manually once?
Phil Hie
ex Goldman Sachs · ex Avelios Medical
March 2026
0
IT tickets analyzed
59%
no owner
49%
automatable
The Evidence
What 2,300 tickets revealed
Every IT support ticket at Avelios Medical (180 employees), Jan 2025 – Mar 2026.
59%
had no owner
No assignment. No SLA. No accountability.
49%
automatable
Password resets, licenses, access provisioning.
50%
handled by one person
Single point of failure.
The real question isn’t how to resolve these tickets faster. It’s why they aren’t automated in the first place.
The Real Problem
Automation tools exist. Nobody uses them.
There’s no shortage of workflow builders. The problem is the friction of building.
1
Describe the automation
“Reset MFA for locked-out users, verify identity, log the action.”
1 sentence
2
Build it
Drag-and-drop in Okta Workflows, ServiceNow Flow Designer, or Zapier. Test. Debug. Maintain.
2–4 weeks
3
You do it manually. Every. Single. Time.
∞
repetitions
“The tool is not the problem. The rules of the game are the problem.”
— IT Governance Lead, Energy Sector
Faster to automate forever than to do it once.
Act 1 — IT Admin builds an automation
Autonomous IT — Builder
MK
Marc K. IT Admin · 9:15 AM
Platform
Building workflow
Validate Role → Check engineering team membership
Check Policy → Auto-approve if engineering, else manager approval
Provision License → JetBrains API, max 2 per person
Notify User → Slack DM with license details
Deploy Automation
Deployed. This automation will now handle all WebStorm license requests.
Build time: 30 seconds
Traditional workflow builder: 2–4 weeks
Now watch what happens when an employee needs a license
Act 2 — Employee triggers the automation
# it-support
SM
Sarah M. 10:32 AM
Autonomous IT 10:32 AM
Checking role and permissions
Role: Engineering → Auto-approved
License count: 1 of 2 → Within limit
JetBrains API → WebStorm license provisioned
Slack DM sent to Sarah
Done. WebStorm has been added to your JetBrains account.
12 seconds total time
Human involvement: Zero
Audit trail: Complete
Integration
Works with your tools. Replaces none of them.
Auto-Resolution
Audit Trail
Compliance
Analytics
Autonomous IT — Governance + AI Layer
Your rules. Your compliance standards. AI builds and executes.
Automation Builder
Policy Engine
Human-in-the-Loop
Audit Trail
Jira
Slack
Teams
Email
Okta
Active Directory
Observe
Week 1
Read-only. See what AI would do.
Govern
Month 1
Rules live. Team retains veto.
Automate
Month 3
Build in natural language. Auto-resolve.
Scale
Month 6+
Mirror data. Go primary when ready.
No forced migration. Ever.
Why SWISSto12. Why Now.
Four reasons this is the right moment.
200+
Hypergrowth
From 6 to 200+ employees, 52% YoY. What worked at 50 breaks at 200.
ESA
Regulated
Defense, cybersecurity, space. Every IT process must be auditable. Governance isn’t optional.
€200M+
Backlog at Stake
HummingSat 2027. Your engineers can’t afford IT bottlenecks during the most critical delivery phase.
EU
Sovereign by Design
European hosting. GDPR-native. No US Cloud Act exposure. DORA and NIS2 ready.
I’d value 15 minutes of your perspective.
We’re building this right now. The ticket analysis was the starting point. Conversations with CIOs confirmed the pattern. Now I’m looking for design partners who can shape the product from day one.
1
What would your IT team automate first if there were zero friction?
2
How does SWISSto12 handle the tension between speed and compliance as you scale toward HummingSat?
3
What would need to be true for you to trust an AI layer on top of your existing service tools?