KPI cards. Network health. The access-network map. Live per-IP traffic. Cashflow and the till. Your task queue. Break scheduling. Live van positions. The Aelita AI assistant. Real-time. Multi-tenant. The COO's morning brief, generated at the moment of opening.
Most ERPs ship "executive dashboards" that pull last night's snapshot from a data warehouse. By the time leadership opens the dashboard at 9am, the numbers are already 12 hours stale. A churn spike that started at 2am goes unnoticed until the next refresh cycle. A fibre cut affecting 200 customers shows zero impact on a customer-facing metric until a batch job runs at midnight.
The ISPCQ dashboard is the operational data spine, queried directly. KPIs reflect what happened in the last 60 seconds, not what happened yesterday. When a contract is created, the count increments. When a payment lands, the balance shifts. When the NOC flags an outage, the affected-customer total updates within the page lifecycle.
Reports that lag get questioned. Reports that update get used.
Nine tabs — Stats, Customers, Network Health, Infrastructure, sFlow, CashBox, My Tasks, Breaks, Vehicles — plus the Aelita AI assistant reachable from any screen. Every tab is gated by permission, so a cashier, a support technician and an executive open the same page and each sees a different dashboard. The CEO doesn't need the cashbox drill-down; the cashbox supervisor doesn't need MRR.
The C-level read of the business, laid out as numbered tiles that fit a single screen and scale down to a phone. Monthly recurring revenue, collections progress and payment-method mix; the live customer base by contract state; bandwidth with a free-versus-paid split; income and expense by class with bill-cycle drill-downs; and the bi-monthly VAT panel. Most figures refresh every 30 seconds; the bandwidth chart every three. Tiles you can't see are hidden, so the cockpit fits your role.
The front-desk surface. Search any customer by account number, address, IP, e-mail, name, or VAT / company registration number, and open their full profile. Beside the search bar, a Live Calls widget shows current and recent phone calls auto-matched to the caller's account — with a coloured bullet when that caller is inside a known outage or planned-maintenance window, so the agent knows before they pick up. Log every in-person arrival in the Walk-in Visitor Log, and ring up a counter sale with Quick Sale — no account needed. Accounts that are both a customer and a reseller now show all their contracts — owned and resold — in one unified, approved-first view, each row labelled Customer or Under Reseller. Every contract card carries a renewal timeline that separates what is running today from renewals scheduled to take over and past terms already terminated, so a glance tells the agent exactly where the service stands. The profile also keeps a per-contract self-repair journal — a read-only history of the automatic repair attempts the platform ran on the connection — so the record stays trustworthy when a customer calls back about a recurring fault. The profile's Journals card turns its Calls tab into a speaker-labelled transcript of every recorded call — click any timestamp, such as 2:14, and the recording jumps straight to that moment; an unanswered after-hours call sits flagged with an amber voicemail badge showing the opening words of the message, so staff see what it was about before they ever press play. A Print Agreement button next to each contract opens the signed customer agreement in a ready-to-print popup, on demand, without leaving the page. For a contract with Advanced Port Configuration — an ONU wired with more than one LAN port set up differently, such as a multi-port router or a free-WiFi hotspot carved onto its own port — the MACs tab tags every unassigned device with the physical LAN port it was heard on, so the agent assigns each device to the right port instead of guessing. Staff whose role includes the redesigned Customers page get a summary band (line status, balance, contracts, open items) above a contract index grouped by what needs attention, with the selected contract beside it and a collapsible side rail of customer, billing and journal cards; a customer with one contract opens straight onto it, and staff without it keep the page described above. On a fibre contract a live ONU chip in the contract header and a live ONU card in Contract Details show the unit's state with a running timer, and whoever presses Reboot is told when the unit is back. Each call card names the agent answering the extension whenever that is certain; where two customers share a phone number, the agent can log the call on either holder and it follows the note to that customer's history.
The support-and-NOC triage board, filtered to the project you're watching and drawn as a live board: counters and a status lamp in the header, a Live or Polling indicator, and an alert strip for access nodes that have stopped responding. It surfaces the things that turn into tickets: how many subscriber units have lost signal over the last 4 hours, 24 hours and 7 days, with the customers behind a node that has stopped responding counted in; which access-line ports have wire-downs in the last three days, worst first, with the affected streets; which connections are running on bad signal; and which access nodes have stopped responding and how many customers each one carries. Open faults, the signal-check queue and a 48-hour chart of wire-down, power-off and tolerance events round it out. The server checks every few seconds which parts changed and the page re-fetches only those, so a long list keeps its scroll position; without live push it falls back to polling every 30 seconds. Customer names and numbers open that customer in the Customers tab while the board keeps running behind it.
The access network laid out geographically, with every node drawn as a colour-coded health marker so a glance tells you where things are well and where they're not. A KPI summary across the top counts nodes, connected subscriber units, weak-signal connections and active outages. The map refreshes itself in the background without resetting your pan or zoom, so you can leave it open on a wall display and trust what it shows.
A bandwidth-investigation tool for support managers and executives. Enter a customer's IP address and watch their live traffic on a rolling chart — inbound and outbound bandwidth, the dominant service, and a full per-flow table with remote host, network, country, port and direction. Group the view by service, tier, application, network, country or peer: tier mode splits traffic into the same cost buckets billing uses (on-net, local, peering, international), while application mode names the well-known services behind it (Netflix, Google, Microsoft, Meta, Cloudflare, Steam, and more). The search only accepts IPs that belong to a real customer, so the tool stays a support instrument, not a network-browsing free-for-all.
A cashier sees their own assigned till — open and close the drawer, accept payments against bundles, individual bills, virtual bills and installment plans, and tick the bills to settle with a single green Pay button. The total shows its currency and how many items are ticked, Select All leaves overdue must-pay bills ticked so the total on screen is what is paid, and a cash payment with no open register warns before the confirmation step (EFT never needs one); if the server does not answer after a payment is confirmed, the cashier is told the result is unknown and to check Last 10 Bills before trying again. On the redesigned Customers page the cashbox opens as its own titled card, with the button reading Close Cashbox while it is open, and an optional Mission Control look, given by role or to one cashier first, shows one row per bill with a status lamp, a Pay button that names the method and a chip for the register's state. Quick Sale rings up counter purchases for buyers with no account. Need to charge for a router, an ONT swap or extra on-site work that isn't part of the monthly fee? An optional one-off button drops a single extra line straight onto the customer account, settle-able online or at the counter. Need a fresh copy of the signed agreement at the counter? The same Print Agreement button available on the customer profile also sits here, in the Individual Contract Payments section, so a cashier can print one on demand without leaving the till. Supervisors and admins instead see every till at once: total in, total out, net balance and expected cash, in dual currency, with the queue of storno and credit-note reversals waiting for two-factor approval. Drill into any session by till, officer or date.
Everything assigned to you, pulled from across every kind of note — contract, customer, internal and lead — so the work finds you instead of the other way round. Headline cards count what's assigned to you, what's urgent, overdue, on your team, escalated, and how you're performing; empty cards hide themselves. Below them, your top tasks sit ranked by priority with category, a preview, the owning team, comment count and status. An orange badge on the tab carries your open count, and any card or row jumps straight into the full task view. The unread notification count now mirrors into the browser tab title (Gmail/Slack-style), so staff catch new items even from another tab.
A shift-floor scheduling board for staff breaks. Your personal card shows today's breaks as colour-coded pills — scheduled, on break now, completed or skipped — with Start and End buttons that appear when it's your turn. Team leads see a timeline with one row per person against a time axis marked with the allowed break windows, for one team or several at once, and can step day by day. Where fingerprint punches are recorded, managers, team leads and HR also get a presence strip that filters the rows, with clock-in and clock-out beside each name; the times are for display and do not change attendance or payroll. Where they hold the right permission, they assign a break by clicking an empty slot, drag a block to move it (snapping to fifteen minutes), and remove one with a click. Past days are read-only.
Every company van carrying a GPS tracker, on a live map inside the platform — read from the operator's own Traccar GPS server, so the monitoring team never logs into a separate tracking system. New positions are pushed to the page within seconds of a tracker reporting; each van shows its type, heading, speed and how long ago it last reported, in the same cockpit look as the Stats tab. Clickable Moving, Parked and Offline counters, a find-a-van box and one button per team narrow the fleet, the tab remembers your filters, map style and view per browser, and staff see only their own teams' vans unless they may see them all. Open any van for its route over a day or any period up to 31 days: the route coloured by speed, stops named after saved places or the van's home, a trips timeline with distance and top speed, a replay that plays the drive back at 1×, 4× or 16×, and Excel or CSV downloads. The driver's name comes from the van-key sign-out and is shown only to those allowed to see it. A van can belong permanently to one agent, with a home location: if it leaves home outside that agent's rostered shift, a task is raised for the monitoring team, and upcoming services, licence renewals and insurance raise reminders of their own.
Aelita is the built-in AI help assistant, one click away from the sidebar on any screen. Ask it about billing, how a module works, how to perform a task, or company policy, and it answers from the platform's own knowledge base — following the conversation so you can keep digging. It knows which screen you're on and uses that as context. Aelita guides and explains; it never makes changes on your behalf, and every conversation is logged for quality review. The same intelligence powers an instant customer-history summary on the profile and a suggested category when you write a note.
The question. 9:15 AM. The COO has heard from the NOC manager that there was a fibre cut overnight. She walks into the office and asks: how many customers were affected, did billing recognise it, are we going to give credit notes, and what's the support backlog look like?
What used to happen. Three teams scramble. Network ops queries the OLT for affected ONUs. Support filters tickets by date. Finance looks at credit-note requests. Three spreadsheets get emailed by 11:30. By the time they're consolidated, it's lunch.
What ISPCQ does. The COO opens the dashboard. The Network Health tab shows last night's incident: 47 affected customers, resolved at 8:30 PM, average outage duration 2h 14m. On credit notes, there's no auto-trigger to check — outage compensation is always a manual, case-by-case decision, so she flags it to a manager on the spot rather than waiting on the system. The Customers tab's Live Calls widget shows the related call volume peaked at 19:20 and tapered by 21:00, each call already flagged with the outage bullet. She types one affected customer's IP into sFlow and confirms the line is healthy again now. One screen, ten seconds, every team's view.