Support
engineering
SIP LAN PoE DECT
I look after business phone systems and the networks underneath them. When a call drops, a handset refuses to provision, or a site loses a VLAN somewhere between the ONT and the desk, I find it — usually in the part nobody wrote down.
London, remote UK & EU hours Contract or retained
01 — Kit
Not a logo wall. These are the systems I have racked, provisioned, broken, fixed and migrated — and whose release notes I read before anyone asks me to.
P-Series and S-Series, TA and TG gateways, Linkus on desktop and mobile. SIP trunk cutovers with the old lines still up, call routing and IVR trees a receptionist can follow without a manual.
Nebula-managed estates, USG FLEX firewall policies and site-to-site VPNs, GS-series stacks, WAX access points. Port profiles, PoE budgets and a QoS policy that actually holds under load.
Omada controllers on-prem or hosted, EAP access points, JetStream switching. VLAN design that keeps the guest network well away from the voice one, and roaming that survives a warehouse.
D-Series provisioning by redirection, M700 bases and multicell DECT that holds a handover across a full floor. Firmware rollouts staged so a bad build never takes the whole site with it.
T5-Series on the desk, W-Series on the move, MeetingBar in the rooms. Zero-touch through RPS and device management, plus the Teams-certified builds and the licensing that comes with them.
02 — Method
In that order, every time. Skipping a step is how a fix lasts a fortnight instead of a year.
Get the fault on demand. A report of "calls sometimes drop" is a starting point, not a symptom — I want the extension, the time, and the leg it died on.
Split the path and halve it again. Handset, switch port, PBX, trunk, carrier. A packet capture settles arguments faster than a conference call does.
Change one thing, with a way back. Config saved off-box before I touch it, and a window agreed with whoever answers the phones.
Test the original fault and the three things nearest to it. Then leave it running long enough to be sure the fault has actually gone.
What broke, why, what changed, what to watch. Handed over so the next person — including future me — doesn't start from zero.
Most outages are a reasonable decision someone made two years ago.
03 — Contact
Send the site, the symptom and roughly when it started. If it's live and business-critical, say so in the subject line and I'll come back inside the hour during working hours.
hugo@fmandrake.duckdns.orgAvailability
Taking on new estates from September. Ad-hoc escalation cover year-round.
Working style
Remote first, on site for cutovers, surveys and anything that needs a hand on the rack.
Also fluent in
FreePBX and 3CX, Draytek and MikroTik edges, and inherited systems with no documentation.