Infrastructure
Network, servers, telephony, peripherals - the ground your IT stands on
The foundation of your IT. Planned properly, run reliably, documented completely - and repairable in an emergency, even when we are not on site.
Was wir in der Säule Infrastructure machen.
3 Topics, jedes einzeln buchbar oder als Gesamt-Setup. Klick auf eine Zeile öffnet die Detail-Page.
| No | Topic | Type |
|---|---|---|
| 01 | Network· LAN, WLAN, VLANs, firewall - planned and documented | Topic |
| 02 | Servers & virtualisation· Proxmox, Windows Server, hybrid - without a licence trap | Topic |
| 03 | Telephony· SIP, VoIP, Teams calling - hybrid and reliable | Topic |
So bauen wir.
Vier Punkte, die in jedem Projekt der Säule Infrastructure unverhandelbar sind.
About this pillar.
What “infrastructure” actually means here
Infrastructure is the thing nobody praises when it works and everybody complains about when it breaks. We build it so the second case happens as rarely as possible, and so the first case still feels solid.
Network
For us, network means everything from the access point in the hallway to the VPN bridge to Berlin. We plan LAN, WLAN, VLANs and firewall rules so that your team never notices any of it - it simply works.
What we typically build:
- Structured LAN with clear VLANs (staff, guests, management, VoIP, IoT)
- WLAN with seamless roaming, WPA3, separate guest networks and zero-trust options
- Firewalls (UniFi UDM Pro, OPNsense, pfSense or Fortinet depending on requirements) with a documented rule set
- VPN (WireGuard for performance, IKEv2 where it is required, zero-trust tunnels via Cloudflare Access as the modern alternative)
- Segmentation - a compromised IoT device must never be able to reach the accounting machine
Servers and virtualisation
We run servers on-prem, in the Hetzner data centre in Germany, and in Azure or AWS. Most often we virtualise with Proxmox - open source, no VMware licence trap, with snapshots, live migration and ZFS storage.
Typical setups:
- Proxmox cluster with two or three nodes and Ceph storage
- Single host for small operations (a NUC form factor is often enough)
- Hybrid: critical services redundant, test workloads on a single host
- Windows Server where it has to be (AD, specific industry software)
Telephony
Telephony sounds dated - but it is often the first thing people notice in a crisis. We deploy classic SIP trunks where that is what you want, and Microsoft Teams Phone where it fits. For medical practices, law firms and trades that do not want to give up desk phones, we build hybrid setups with SNOM or Yealink handsets plus a cloud PBX.
Hardware lifecycle
Laptops, monitors, printers, peripherals. We procure, roll out, maintain and dispose. No storage costs on your side, no forgotten MacBooks in drawers.
How our projects usually run
- Intro call - 45 minutes, free. You talk, we listen.
- Inventory - one to three days on site. We document what exists (network topology, servers, identities, backup, monitoring, open issues).
- Concept - in writing, with reasoning. You decide what gets built.
- Build - with a schedule and regular stand-ups. No surprises.
- Operations - monthly flat rate or time and materials. The documentation stays with you - even if we part ways one day.
How we work together.
From the first hello to running operations in five steps.
Infrastructure - frequently asked.
FAQ.01What counts as "infrastructure" the way you use the word?
Everything below the application layer - network (LAN, WLAN, VPN), servers (on-prem and cloud), storage, telephony (including Teams calling), endpoints (clients, printers, peripherals), monitoring and backup hardware. In short: the things that have to work before anything else can.
FAQ.02Can you plan a network that we then build ourselves?
Yes. Planning and documentation are a service in their own right, separate from operations. You get an implementation package with parts lists, configurations and wiring diagrams - your electrician builds it. After that we either commission it or you do it yourselves.
FAQ.03Which network vendors do you recommend?
Mostly UniFi (Ubiquiti) for small and mid-sized setups, and Arista or Juniper where the requirements are higher. The choice depends on your budget, the expected service life and how much central management you need. We only deploy Cisco Meraki when a client specifically asks for it - we rarely find the licence costs justified.
FAQ.04Do you do the cabling yourselves?
Planning yes, pulling cable no. We work with established electrical contractors in the Rhine-Main region who build our plans properly. You get one invoice, we handle the coordination.
FAQ.05On-prem servers or cloud?
Both - and often hybrid. We recommend cloud where it makes sense (mail, collaboration, ordinary office workloads) and on-prem or hosted where latency, data protection or cost argue against it (smart-building gateways, media assets, AI workloads running local models). Our rule of thumb: the more specific the workload, the stronger the case for on-prem.
FAQ.06How does the handover from our current IT provider work?
We start with a takeover inventory (one to two days on site, usually plus remote work), document what exists, then take over step by step. A short period of parallel operation is possible - but we insist on clarity about who is responsible from which date. No "we will take over at some point" - a fixed handover date instead.
Sprechen wir 30 Minuten über Infrastructure.
The intro call is free. No sales pressure. We listen, check whether we fit - and tell you straight.