Gábor Temesi Lead Service Engineer · Telecommunications Hungary

Changing the network underneath services that can’t stop.

Ten years of hands-on work across fiber, routing, WAN architecture, security, voice systems, virtualization and automation — applied to migrations of 100+ location networks, satellite-connected voice and 24/7 call centers.

Get in touch Read the case studies
Call-center relocation window
<48hto move a 24/7 call center — no outage, no lost calls
SD-WAN migration scale
100+locations moved from DMVPN to Cisco Catalyst SD-WAN
Largest access-technology rollouts
150+locations in the largest access-technology rollouts
Experience
10yrsof field and engineering work in telecommunications

01 / Profile

An engineer who can follow a problem the whole way down.

I’m a lead service engineer in telecommunications. My work runs from the physical layer — copper, fiber and wireless — through routing, WAN architecture and network security, up to the servers, virtualization platforms and voice systems that run on top.

That range comes from doing the work rather than reading about it. I’ve spliced and measured fiber, installed equipment on site and debugged SIP at packet level, and I plan and deliver network migrations across distributed, business-critical environments.

When an issue crosses the line between network, system and application, I don’t have to hand it off at the boundary. And where it helps, I write the code — Bash, PHP and others — to integrate components and automate repetitive operational work.

Role
Lead service engineer, telecommunications
Experience
About ten years, in the field and in the office
Focus
WAN migrations, mission-critical voice, infrastructure
Based in
Hungary
Languages
Hungarian (native), English (professional working proficiency)

02 / Selected work

Engineering work in live environments.

Real engagements from my professional work. Employer and customer details are deliberately left out. The diagrams are simplified conceptual illustrations drawn for this site, not the deployed architectures.

Case 01Service continuity

Relocating a 24/7 call center in a single weekend

The complete, operational infrastructure of a round-the-clock call center — external lines, telephony, network, security and servers — physically moved to a new location in under 48 hours, while it kept taking calls.

< 48 hmaintenance window

0 outagesservice stayed up

0 lost callsnot a single one

Case study breakdown · context, challenge, role, scope, outcome
Context
A call center operating 24/7, with its entire infrastructure on one site that had to move to a new location.
Challenge
The service could not stop. Every layer — external connectivity, telephony, network, security and compute — had to be moved, reconnected and back in service within one weekend, with less than 48 hours available.
My role
Planning, preparation and execution of the relocation.
Scope
  • Leased lines
  • Telephone lines
  • PBX & telephony
  • Network equipment
  • Firewalls
  • Servers
  • Rack cabinets
Outcome
The complete infrastructure was relocated with no service outage and without losing a single call.
Relocation within a weekend window Leased lines, telephone lines, PBX and telephony, network, firewalls and servers move from the original site to the new site within a window of under 48 hours, while the call service line runs continuously underneath. WEEKEND WINDOW · UNDER 48 H ORIGINAL SITE Leased lines Telephone lines PBX & telephony Network Firewalls Servers relocate NEW SITE Leased lines Telephone lines PBX & telephony Network Firewalls Servers CALL SERVICE continuous no outage · 0 calls lost
Conceptual illustration. It shows what moved and the result; it does not describe the procedure used.

Case 02WAN transformation

DMVPN to Cisco Catalyst SD-WAN across 100+ locations

Planning, preparing and executing the migration of an existing DMVPN network with more than 100 locations to Cisco Catalyst SD-WAN, in a design that departs from the usual reference model.

  • Controllers and hub routers hosted on premises, not in the cloud
  • Hub-and-spoke topology: spokes communicate only with the hub, not with each other
  • Underlay over private MPLS VPNs, not the public internet
  • The technology learned and applied during the deployment itself
  • Configurations and the rollout workflow refined as real issues surfaced
Case study breakdown · context, challenge, role, scope, outcome
Context
A distributed network of more than 100 locations built on a DMVPN architecture, moving to Cisco Catalyst SD-WAN.
Challenge
The target design didn’t follow the common pattern of cloud-hosted controllers and an internet underlay. Controllers and hub routers ran on premises, the underlay was private MPLS VPN, and the overlay was hub-and-spoke rather than a full mesh. The technology was also new to me, so a working understanding had to be built during the deployment itself.
My role
Planning, preparation and execution of the migration; troubleshooting the real-world issues that appeared during rollout; iteratively optimizing both the configurations and the deployment process.
Scope
  • Cisco Catalyst SD-WAN
  • On-prem controllers & hubs
  • Hub-and-spoke overlay
  • DMVPN
  • MPLS VPN underlay
  • Overlay networking
Outcome
A 100+ location network migrated from DMVPN to Catalyst SD-WAN, with a deployment workflow that improved as the rollout progressed.
DMVPN to Catalyst SD-WAN, conceptually Before: sites connect to an on-premises hub through DMVPN tunnels. After: the SD-WAN Manager, Controller and Validator and the hub routers are all on premises. Spokes connect through the SD-WAN overlay to the hub routers only, with no direct spoke-to-spoke traffic, over a private MPLS VPN underlay. BEFORE · DMVPN hub on-premises site site site site site hub-and-spoke tunnels 100+ locations AFTER · CATALYST SD-WAN on-premises Manager Controller Validator hub routers spoke spoke spoke spoke spoke spokes reach the hub only · no spoke-to-spoke UNDERLAY · PRIVATE MPLS VPN
Conceptual illustration. Dashed: DMVPN tunnels. Orange: SD-WAN overlay, hub-and-spoke only. Controller connections to the edges are omitted for clarity; site counts are simplified.

Case 04Custom telephony engineering

A redundant, load-balanced call-routing layer built on Asterisk

Designing and implementing a custom Asterisk-based solution that serves an SBC-like role for complex call-routing scenarios: two nodes, load-balanced on both legs, with automatic failover and TLS-secured connections to cloud services.

  • Two Asterisk nodes, load-balanced on both the upstream and the internal legs
  • Trunk states monitored on both legs
  • If a node becomes unreachable, all calls are routed to the surviving node
  • SIP over TCP and TLS, with publicly trusted certificates as required for cloud connections
  • Complex call-routing logic implemented in software
Case study breakdown · context, challenge, role, scope, design
Context
Complex call-routing scenarios that called for a dedicated, resilient routing layer.
Challenge
Combining complex routing logic with redundancy, load balancing and secure cloud connectivity in a custom-built system rather than an off-the-shelf product.
My role
Design and implementation of the solution.
Scope
  • Asterisk
  • SIP/TCP
  • SIP/TLS
  • Public certificates
  • Call routing
  • Load balancing
  • Failover
  • Trunk monitoring
Design
Two Asterisk nodes. Both the upstream and the internal legs are load-balanced, with trunk states monitored. If the connection to one node is lost, all calls are routed to the node that is still available. Signalling runs over SIP/TCP and SIP/TLS, using valid public certificates as required for the cloud connections.
A precise note
This is an Asterisk-based implementation performing SBC-like functions — not a commercial, standards-certified session border controller product.
Two-node Asterisk routing layer, simplified Upstream voice connectivity is load-balanced across two Asterisk nodes, which connect to internal telephony platforms. Trunk states are monitored on both legs, and if one node becomes unreachable all calls are routed to the other. Signalling uses SIP over TCP and TLS. ASTERISK-BASED · SBC-LIKE ROLE SIP/TCP · SIP/TLS UPSTREAM VOICE CONNECTIVITY load-balanced trunk monitoring Asterisk node 1 Asterisk node 2 failover load-balanced trunk monitoring INTERNAL TELEPHONY PLATFORMS
Simplified illustration of the two-node design. Orange: if one node becomes unreachable, calls move to the other.

Further engagements

Rollouts · hybrid connectivity · dispatch systems

Case 05Network transformation

Access-technology migrations across networks of up to 150+ locations

Planning, preparing and executing technology and device rollouts that replaced legacy connectivity across geographically distributed networks — choosing the right access technology per site, keeping configurations consistent across many locations and moving live services onto the new connections.

  • ISDNreplaced by 4G/5G, xDSL or other suitable access
  • xDSL & Multiflexreplaced by high-speed leased lines over private MPLS VPN
  • Legacy point-to-pointreplaced by DMVPN-based architecture

Case 06Hybrid connectivity

The network side of a call center’s move to the cloud

A complete call-center environment moved from on-premises infrastructure to a cloud-based solution. My part was the network: designing and implementing the new connections and the infrastructure changes the transition depended on.

For business-critical communications, the network path between the existing environment and the cloud service is part of the service itself — it has to be in place and right before anything can move.

  • Connectivity design
  • Hybrid infrastructure
  • Service integration
  • Call-center networking

Case 07Mission-critical voice

Operating Mitel OpenScape Xpert dispatching systems

Operation and maintenance of multiple Mitel OpenScape Xpert dispatching infrastructures — specialized voice communication systems that need continuous operational attention and an understanding of the network and systems they run on.

  • Mitel OpenScape Xpert
  • Dispatch communications
  • Operations & maintenance

03 / Expertise

Seven domains, one connected skill set.

Each area below is something I’ve worked with directly. Together they cover the path a service actually takes — from the cable in the wall to the application on top.

  1. Routing, switching & WAN

    Designing, configuring and troubleshooting enterprise routed and switched networks and the WANs that connect distributed sites, across multiple vendors. Routing and switching knowledge at Cisco CCNP and MikroTik MTCINE level.

    Platforms
    Cisco, Juniper, MikroTik
    Protocols
    OSPF, EIGRP, BGP, BFD, MPLS, VLANs, STP, link aggregation
    WAN
    Cisco Catalyst SD-WAN, DMVPN, MPLS VPN, leased lines, xDSL, 4G/5G, Starlink

  2. Physical & telecom infrastructure

    Planning, building and maintaining the passive layer: structured cabling, fiber-optic links and wireless installations, including on-site installation and fault finding.

    Copper
    Structured cabling, CAT3 to CAT7
    Fiber
    Splicing, repair and measurement
    Other
    Wireless installations, passive telecom infrastructure

  3. Voice & real-time communications

    Enterprise telephony, dispatching systems and the protocols underneath them — down to packet-level SIP and RTP debugging and the TLS layer that secures them.

    Systems
    Mitel OpenScape 4000, Mitel OpenScape Xpert, Asterisk
    Function
    PBX, SBCs and softswitches, multicast voice
    Protocols
    SIP, RTP, SIP over TLS, SRTP

  4. Security, access control & PKI

    Operating perimeter security and network access control, centralizing authentication, and building secure site-to-site connectivity.

    Firewalls
    NGFWs: URL filtering, IPS, advanced ACLs, GeoIP filtering
    Access
    Cisco ISE, RADIUS, LDAP
    Crypto
    IPsec, WireGuard, site-to-site VPN, PKI and certificate troubleshooting

  5. Systems & virtualization

    Running the servers, hypervisors, storage and backup that infrastructure services depend on — from hardware to identity and supporting services.

    Systems
    Linux, Windows
    Virtualization
    Proxmox VE, Proxmox Backup Server, VMware ESXi and vCenter, VMs and containers
    Hardware
    Dell and HP servers, EMC and IBM storage, tape libraries

  6. Automation & scripting

    Writing code to integrate infrastructure components and remove repetitive, error-prone operational work — the bridge between infrastructure and software.

    Languages
    Bash, PHP, VBScript, Java, C#
    Tooling
    Ansible (basic)

  7. Monitoring, operations & troubleshooting

    Keeping infrastructure observable and recoverable, and finding root causes when symptoms and origins sit in different layers.

    Monitoring
    Nagios, Zabbix, LibreNMS
    Operations
    RANCID configuration backup, reverse proxies
    Method
    Packet-level analysis, cross-layer root-cause analysis

Depth varies by area, as it does for any engineer. The case studies above show where it runs deepest: WAN architecture, migrations and voice.

04 / Approach

How I work.

See the whole system

Symptoms appear in one layer and originate in another. I trace a problem across the physical link, the network, the system and the application before deciding where the fix belongs.

Fix the cause, not the symptom

A workaround is sometimes the right call for tonight. It shouldn’t become the permanent answer, so I follow through to the root cause.

Prepare more than you improvise

Changes in tight windows are won before they start. Thorough preparation is what makes a short maintenance window survivable.

Keep it understandable

A design the next engineer can read and operate is worth more than a clever one only its author understands.

Automate what repeats

Scripts earn their place by removing repetitive, error-prone work — especially across many sites whose configurations have to stay consistent.

Balance reliability, security and simplicity

Each matters, and each can be overdone. The goal is a system that is secure enough, resilient enough and still simple to run.

05 / Background

Both sides of the work.

About ten years in telecommunications, now as a lead service engineer. The experience comes from two directions: on site with the hardware, and at the desk with the design.

In the field

Hands on the infrastructure

  • Installing and maintaining copper, fiber-optic and wireless networks
  • Fiber splicing, repair and measurement
  • Structured cabling from CAT3 to CAT7
  • On-site installation and troubleshooting of network and telecom equipment
  • Server hardware, storage systems and tape libraries

In the office

Planning, design and operations

  • Planning and preparing technology rollouts and migrations
  • Designing and configuring routed, switched and WAN networks
  • Operating NGFWs, Cisco ISE and SD-WAN environments
  • Running virtualization, backup and infrastructure services
  • Packet-level debugging of voice and network issues

Qualifications

Skills built on the job and through continuous self-study. Routing and switching knowledge at the level of Cisco CCNP and MikroTik MTCINE (Certified Inter-Networking Engineer) — working knowledge and experience, not held certifications.

Outside work

Personal projects with self-hosted services, infrastructure automation, authentication systems and network labs — where new technologies get tested before they’re needed.

06 / Contact

Let’s talk.

For roles, projects or a technical conversation, email is the most direct way to reach me.

Write an email

Based in Hungary

Languages Hungarian (native) · English (professional)