Redundant border control
Purpose-built SIP border controllers deployed in redundant configuration, so signalling survives the loss of any node.
The platform
We run the network ourselves.
Most providers in this market resell someone else's routes behind a web panel. HiClass is the other kind of company: we design, build and operate our own carrier-grade voice and messaging platform — and this page describes the machine itself.
Architecture
Built for carrier traffic.
The core is ours end to end: signalling, media, data and failover are engineered as one system, not assembled from rented parts.
Signalling & media
Traffic enters through purpose-built SIP border controllers running in redundant pairs, with an active-active fleet of media relays behind them. Signalling and media scale independently, and no single machine is a point of failure.
Termination coverage spans a very large prefix set across destinations covering most of the world, managed through our own routing layer — which means routing decisions, quality controls and failover policy are ours to engineer, not a supplier's black box.
Active-active media fleet
Media relays run active-active: capacity is added horizontally and failure is absorbed without dropping established calls.
Clustered data layer
A clustered database with automatic failover, replicated caching, and offsite backups — the platform's state survives hardware, not the other way round.
Wide termination coverage
A very large prefix set spanning destinations across most of the world, curated and quality-controlled by our own routing engine.
Visibility
If it happens on the wire, we can see it.
Operating the platform means owning its instrumentation. Nothing on this list is a bolt-on dashboard — it is how we run the network every day.
Live call monitoring
Active calls are visible as they happen, across the whole platform — not sampled, not delayed.
Immediate call-detail records
CDRs are written as calls complete and are available at once, with per-second granularity feeding billing directly.
SIP-level packet capture
When something needs explaining, we read the actual signalling — full SIP traces, on our own infrastructure, without waiting on a third party.
Metrics & log pipeline
A full metrics and log pipeline across every component, monitored around the clock.
Defence
Fraud is an engineering problem. We treat it like one.
Premium traffic attracts abuse. Our protections are built into the platform layer, where they can actually see the traffic.
IRSF detection
International revenue-share fraud patterns are detected by the platform itself, watching real signalling rather than yesterday's reports.
False-answer supervision
FAS detection guards billing integrity — calls that were never really answered are caught, not invoiced.
Per-provider limits
Rate and concurrency limits are enforced per provider, so one misbehaving source cannot degrade the platform for everyone else.
Traffic-pattern analysis
Automated analysis watches traffic shape continuously and flags what a human should look at.
Billing & self-service
Precise numbers, no surprises.
The commercial layer
Billing runs on the same per-second call detail the platform writes — automated invoicing, wallet and payout cycles, all computed from primary data rather than reconciled afterwards.
Customers work in a self-service portal built on the platform's own APIs: number allocation, IVR upload, test numbers and reporting — with a hierarchical reseller and sub-account structure where each level carries its own independent limits.
- Per-second call detail feeding automated invoicing
- Wallet and payout cycles computed from primary data
- Self-service portal: numbers, IVR, test numbers, reporting
- Hierarchical reseller and sub-account structure with independent limits
Built by the people who run it.
If you want this platform terminating your traffic — or want us to build one like it for you — talk to the engineers who operate it.