AI network management · NOC-as-a-service

Every alert, understood.

Secure is an AI network management system with a human NOC behind it – watching your switches, firewalls, links, servers and LoRaWAN fleets, compressing alert storms into root causes, and telling you what matters in plain English. You own the infrastructure. We mind it.

Born in cyber security · UK-based · for networks you own – from one office to a borough
24/7

the system never sleeps; humans are on call for the moments that matter

1 pane

IT and IoT estates in the same console – no more swivel-chair monitoring

LoRa native

gateways, join health, battery and airtime as first-class signals, not an afterthought

0 lock-in

open exports, your hardware, documented handover – leave whenever you like

The problem

Monitoring that shouts is not monitoring.

Most estates already have alerts. What they don’t have is comprehension – something that reads the storm, names the cause, and knows which three of the hundred and forty messages deserve a human.

// alert fatigue

Hundreds of alerts, three that matter

Every device cries wolf in its own dialect. After the tenth false alarm, people stop reading – and the eleventh is the real one.

// tool sprawl

Your NMS can’t see the sensor fleet

The network tool doesn’t speak LoRa. The LoRaWAN console doesn’t see the WAN. The server tool sees neither. Nobody sees the whole estate.

// no NOC

24/7 eyes are a team you don’t have

A round-the-clock network operations centre is five salaries and a rota. At council and SME scale that maths never works in-house.

// found by users

The phone should not be your monitor

If the first sign of an outage is a complaint – or worse, a security incident found by someone else – the monitoring already failed.

The console

Tuesday, 03:12. Nobody got woken up.

A gateway power supply starts sagging in the small hours. Fourteen sensors go quiet, the LNS floods with join failures, and a calendar-grade monitoring tool would have paged a human fourteen times.

Secure reads the storm as one incident, names the component, checks it against the baseline, and queues the fix for the morning – with the overnight evidence already written up.

This console is a simulation. Your estate is where it stops being one. This screen, your network – book the review
Coverage

One console, the whole estate.

From the firewall to a sensor on a lamp post – if it has a heartbeat, Secure listens to it. Read-only by default, least privilege always.

Switching & Wi-Fi

Port health, PoE budgets, spanning tree surprises, AP client experience and rogue-AP detection.

WAN, VPN & links

Circuit utilisation, latency and jitter baselines, tunnel health, failover that actually fails over.

Firewalls & edge

Rule hygiene, session anomalies, certificate and firmware currency – watched with a security eye.

Servers & services

Host vitals, service reachability, disk and cert expiry forecasts, backup job verification.

Cloud & SaaS

Tenancy health, API quotas, spend anomalies and the dependencies your services quietly stand on.

The speciality

LoRaWAN fleets

Gateways, join health, RF noise, airtime and duty cycle, battery forecasts, device identity and key hygiene – the estate most NMS tools cannot see at all.

The guru

AI that triages. Humans who decide.

The model does the reading, clustering and explaining. Changes to your network are made by engineers, under change control, with an audit trail – the AI proposes, a human disposes.

01 / compress

Storm → root cause

Correlates a flood of device-level alerts into one named incident: not “fourteen sensors down”, but “GW-01 PSU degrading – everything behind it is collateral”.

02 / baseline

Learns normal, flags weird

Thresholds are guesses; baselines are evidence. Secure learns each element’s rhythm and surfaces deviations – including the quiet kind that precede failures.

03 / narrate

Plain-English incidents

Every incident closes with a written narrative: what broke, why, what was done, what to change. Readable by a director, not just an engineer.

04 / predict

Sees failures coming

Battery curves, disk growth, certificate expiry, duty-cycle creep – the predictable failures get scheduled as maintenance instead of arriving as outages.

05 / propose

Suggested runbooks

Each incident arrives with a proposed next action drawn from your estate’s history. Engineers approve, amend or reject – and the system learns from that, too.

06 / suspect

Security-tinted ops

An outage and an intrusion can look identical in minute one. Anomalies are treated as guilty until explained – a habit we brought from cyber security.

The speciality

We speak LoRa natively.

Thousands of battery devices, gateways on rooftops, a network the client actually owns – LoRaWAN estates are wonderful and almost universally unmonitored. Secure treats them as first-class infrastructure: same console, same baselines, same triage as the rest of your network.

Gateway health & backhaul Join-failure analytics Airtime & duty cycle Battery forecasting RF noise floors Device identity & key hygiene Multi-vendor tenancy

The network behind the bins

Secure is the network layer behind BinSense, whose smart-waste deployments run on LoRaWAN networks the council owns – designed, secured and, where wanted, operated by Secure.

That model is the point: the authority keeps the asset, chooses who runs it, and can host any compliant vendor’s sensors on it – bins today, gullies, air quality and damp monitoring tomorrow. No vendor cloud holding your infrastructure hostage.

BinSense and Secure are part of the same group (CIaaS Limited) – we say so plainly.

The service

Three ways to take it.

Start where it hurts. Every tier includes the console, the AI triage and a monthly service report written in plain English – no figures we can’t stand behind, no contract theatre.

Tier 01

Watch

For teams who want eyes, not hands – you fix, we find.

  • Full estate monitoring – IT and IoT in one console
  • AI triage & baselines with plain-English incident narratives
  • Alert routing to your team, your way
  • Monthly service report & review call
Scope Watch
Tier 02

Run

The NOC you don’t have to hire – we find, we fix, you sleep.

  • Everything in Watch, plus hands
  • Incident response under agreed response times
  • Patching, config management & change control with audit trail
  • Vendor liaison – we sit on hold so you don’t
Scope Run
Tier 03

Build

Networks designed to be owned – including council LoRaWAN.

  • Design & commissioning of networks the client owns
  • LoRaWAN coverage surveys, gateway siting & key ceremonies
  • Security architecture from day one, not bolted on
  • Handover documentation – built to be operable by anyone
Scope Build

On pricing: it depends on your estate, so we won’t print a number we’d have to caveat into meaninglessness. The 30-minute review ends with a written scope and a fixed quote – and if Secure isn’t the right fit, we’ll say so.

House rules

Born in cyber security. It shows.

Guilty until explained

Anomalies get investigated like incidents, not snoozed like notifications. Most are innocent. The habit is the point.

Least privilege – even for us

Collectors are read-only by default and connect outbound only. Write access exists where you grant it, logged when we use it.

Evidence, not vibes

Every incident closes with a written narrative and every monthly report shows its workings. If we can’t evidence it, we don’t claim it.

UK-based, UK-hosted

Your telemetry stays in UK regions, and the people reading it are here too – a phone call, not a ticket queue in another timezone.

Questions

Asked by every IT manager. Answered straight.

How does Secure connect to our network?+

A lightweight collector sits inside your estate and connects outbound only – no inbound holes in your firewall. It reads SNMP, flows, APIs and syslog with least-privilege credentials, and for LoRaWAN it hooks into your LNS. Read-only by default; write access only where you grant it for the Run tier, under change control.

Where does our data live – and what do you actually see?+

Telemetry is processed and stored in UK regions, encrypted in transit and at rest. We see device health and traffic metadata, not your payload content. Retention is agreed in the service definition, and our privacy notice covers the rest.

Is the AI making changes to our network?+

No. The AI reads, correlates, explains and proposes. Changes are made by engineers under change control with a full audit trail – and only on the Run tier, only with the access you’ve granted. The day we let a model push config unsupervised is the day we deserve to lose your business.

Do you only do LoRaWAN?+

No – switches, Wi-Fi, firewalls, WAN, servers and cloud are the bread and butter. LoRaWAN is the speciality because almost nobody else monitors it properly, and because our group builds council-owned LoRaWAN networks that deserve better than an unwatched dashboard.

What happens if we leave?+

You take everything: full telemetry export in open formats, your configurations documented, the collector uninstalled cleanly. Networks we build are handed over with documentation written to be operable by anyone. Lock-in is a business model; it isn’t ours.

Who is behind Secure?+

Secure is the network division of CIaaS Limited, a UK group founded from a cyber-security background, and sibling to BinSense (council smart waste) among other products. Where our companies work together on your project, we say so plainly – you’ll never discover the relationship in the small print.

The ask

Thirty minutes. Bring your topology – or just your alert fatigue.

A network review with an engineer, not a salesperson: what you run, what’s watching it today, where the blind spots are. You leave with a written scope and a fixed quote – and if it’s not a fit, we’ll say so.