Engineered for ISPs that can't afford to lose a taka
Billing, accounting and network control are designed so nothing is silently dropped — even during an outage — and so the platform stays fast as you grow past tens of thousands of subscribers.
30,000+
subscribers per ISP
~1,600
tenant databases
Sub-second
period reports at scale
Zero
lost postings or reconnections
Zero-loss accounting queue
If the accounting service is offline, postings aren't lost — each failure retries with exponential backoff (5m → 20m → 1h → 4h → 12h), event-driven so there's no overhead when healthy, and the core transaction is never blocked.
Durable network retry queue
A failed router or RADIUS operation is queued, de-duplicated and retried up to 10 times every 2 minutes — so a power outage never loses a paid customer's reconnection.
Transactional outbox
Business records and their queue message are written in the same database transaction and published with broker confirms, so async work survives an outage with end-to-end correlation.
Built for 30,000+ subscribers
Set-based SQL aggregation, mandatory column projection and batched lookups keep pages fast at scale — a period summary stays sub-second across 36 months of 30,000 customers.
Exactly-once billing
Two-level deduplication locks mean the nightly cron, a manual mass-generate and a new signup can all fire at once and still produce exactly one bill per customer per month.
Real-time, async, scheduled
A three-channel stack — SignalR live dashboards, Hangfire cron and RabbitMQ async — handles real-time, background and bulk work without blocking the operator.
Scale is a design rule here, not an afterthought
See exactly what ISP Digital recovers for your ISP.
Book a 30-minute demo. We'll map your billing, network and accounting onto the platform and show you the numbers — no obligation.