References

Projects that are running in production

A cross-section of our project work: the starting point, what we built and the result. We do not name clients — we do describe the technical substance.

20+

Years of IT practice

Documented projects

2 h

Downtime on the Hyper-V migration

Sites in the DR design

Project team in front of the mobile camera and security setup during on-site acceptance

IP video technologyCentral VMSCamera analyticsVLAN / subnettingAccess control

SecurityMulti-branch businessSeveral sitesDelivered

Central camera system with network segmentation

Starting point

Numerous IP cameras spread across several branches, each installation managed on its own. Recordings were neither centrally accessible nor stored consistently against the required retention periods.

What we built

All cameras were integrated into a central management and recording platform, with permissions per area of responsibility and alarms from camera analytics — line crossing or unauthorised entry, for example. Because of the high video load, a dedicated network was built and camera traffic separated from the corporate network via VLANs and subnets.

Result

Security staff access the cameras they are cleared for from one place, recordings are retained correctly, and video traffic no longer loads the data network. An incident in the camera segment does not spread to other segments.

Server operations and replication across two sites

ReplicationVirtualisationRPO / RTOBackup & archivingFailover scenarios

InfrastructureCorporate ITTwo data centre sitesDelivered

Disaster recovery across two sites

Starting point

Critical applications and data sat at a single site. A fire, a power failure or a hardware fault would have stopped operations — with data loss the business could not have recovered from.

What we built

Critical data and virtual servers are replicated to two sites: the primary site active, the second on standby to take over in an emergency. Recovery scenarios were designed against defined RPO and RTO targets. Backup was planned alongside for archiving, long retention and systems where a longer RPO/RTO is acceptable, scheduled to keep the load on production minimal.

Result

Critical systems keep running at the second site in an emergency. Downtime and the risk of data loss are significantly reduced, and the dependency on a single site is gone.

Server racks with patch panels and fibre cabling during installation

P2V migrationWindows ServerLinuxHypervisorResource planning

InfrastructureCorporate ITOn-premiseDelivered

Physical servers virtualised (P2V)

Starting point

Operations depended on individual physical servers. Backup, restore and hardware replacement were each an intervention that carried downtime risk.

What we built

Inventory of operating system and application dependencies, then P2V migration of the Windows and Linux servers into the virtualisation environment. After migration, CPU, RAM, disk and network were re-sized and services and application access verified.

Result

The hardware dependency is gone. Backup, restore, migration and resource management now run through the virtualisation layer instead of hardware appointments.

Firewall and network hardware in the rack

Firewall clusterHigh availabilityFailover testingLine redundancySite-to-site VPN

SecurityMulti-branch businessHead office and branchesDelivered

Firewall cluster with high availability

Starting point

The link between head office and branches depended on a single firewall and a single internet line. Any fault would have hit every site.

What we built

A redundant firewall cluster of two devices with automatic takeover on failure, plus line redundancy through a second internet connection. Failover scenarios were tested together with the customer: loss of one device, hardware fault, complete loss of one line, interruption of the network link.

Result

In testing the second instance took over each time and the link between head office and branches stayed stable. The dependency on a single firewall or line is gone.

Network components and structured cabling

VLANSubnettingAccess controlNetwork analysis

SecurityCorporate ITGermanyDelivered

Network segmentation in the corporate network

Starting point

Users, servers and devices all sat in the same flat network. Broadcast traffic weighed on operations, and an incident on one device could have spread unchecked.

What we built

Analysis of the existing network, then a split into smaller, manageable segments using VLANs. Dedicated VLANs and subnets were created for user and device groups, assignments set cleanly, and communication between segments limited to what is actually needed.

Result

Access is controlled, broadcast domains are considerably smaller, and a security problem stays inside its segment instead of spreading across the network.

Virtualisation platform during a migration

ProxmoxHyper-VVM migrationRollback planningPost-migration checks

InfrastructureCorporate ITOn-premiseDelivered

Migration from Proxmox to Hyper-V

Starting point

The existing virtualisation had to move to Microsoft Hyper-V — without halting operations for days.

What we built

Analysis of the existing Proxmox environment, then a migration plan covering operating system, disk layout, network configuration and service dependencies. Backups and a rollback scenario were prepared before the switch, the migration carried out to plan, and critical services verified individually afterwards.

Result

The migration was completed in full; total downtime stayed at around two hours. Operating systems, network, disks, services and application access were checked and confirmed after the switch.

Computer vision system for defect detection in production

Computer visionEdge computingIndustrial camerasPython

SoftwareTextile industryProduction lineIn production

Defect detection in textile production

Starting point

Fabric defects were only noticed at final inspection. Until then the line kept running — the scrap had already been produced.

What we built

Computer vision directly at the line on edge hardware: the camera detects defects while production runs and reports them immediately, without image data leaving the plant.

Result

Defects are caught while the fabric is running, not at the end of the shift.

AngularREST APIPayment integrationPostgreSQL

Own productOwn productCloudOwn development

B2B trading platform

Starting point

Dealer pricing, approvals and payment routes were handled by phone, email and spreadsheets.

What we built

Our own trading platform with customer-specific pricing, dealer accounts, approval workflows and several payment providers connected.

Result

Orders run through one path instead of three channels. Prices and approvals are recorded and traceable.

Similar starting point?

Briefly describe what you run and where it gets stuck. We reply within 24 hours on working days with an initial assessment.

Request an initial call