2FA login. Online payments. Plan upgrades. ONU reboot. Portal Badger self-repair. Customers solve their own simple problems; staff get the call volume they actually need to handle.
The cost of an ISP support call is not the agent's salary; it is the cumulative cost of every call that should not have been made. Customer pays online, agent picks up the phone to confirm the payment was applied. Customer wants to upgrade their plan, agent reads back the same plan options the customer already saw on the website. Customer's WiFi router is misconfigured, agent walks them through a reboot.
None of these conversations need a human on either end.
The Customer Portal handles them. Online payment with instant ledger update. Plan changes with proration applied automatically. ONU reboot with one tap. Portal Badger walks a customer through self-repair for the most common connectivity issues, only escalating to a human when the diagnosis says a human is genuinely needed.
The result is not "fewer customers"; it is "the calls that arrive deserve a human."
Fourteen self-service flows shipped today. Each one took thousands of "could the customer just do this themselves?" support tickets to refine.
Two-factor authentication via SMS or email. Mandatory for accounts above a configurable balance or service tier. Account recovery flows that don't require a phone call. Session management with device-list visibility.
Customers see their open bills, select which to pay (or pay all), and complete checkout through the operator's chosen payment gateway. Stripe and iKhokha are both wired and ready; the operator switches gateways from configuration without touching code. Deposits for future bills supported. Payment lands on the ledger before the customer closes the browser.
Customers see eligible plans (filtered by their address coverage and contract terms), preview the proration impact, and confirm. Pause requests follow the same flow with a reason field. All changes carry the same proration math the staff billing-engine uses. If a service was paused for an overdue balance, the customer doesn't need to call in once they've paid up — the portal recognizes the balance is settled and lets them resume the service themselves.
Portal Badger walks the customer through a self-diagnosis: ONU signal check, OLT port status, recent reboot history, common WiFi misconfiguration. If the issue can be resolved by an ONU reboot, the customer taps once; the system reboots the ONU and verifies recovery. If diagnosis says a human is needed, escalation is one button. The customer doesn't waste time on an avoidable phone call; staff don't waste time on issues a reboot fixes.
Customers open tickets directly through the portal with screenshots, photos, and document attachments. Tickets land in the same queue as agent-created ones, with the same workflow, escalation matrix, and Aelita context-assembly applied.
For ISPs that work through resellers and channel partners, the reseller portal exposes a partner-managed view of just their customer portfolio. Resellers handle billing, support, and plan management for their own customers without seeing other resellers' or direct customers' data.
Happy customers are an ISP's cheapest acquisition channel. The portal turns that into a self-running referral program: customers refer others, and every successful sign-up earns them account credit. They watch it all from their own dashboard — total credits earned, this month's earnings, friends referred, and which referrals are still pending — with a full credit history they can filter by period. No spreadsheets, no manual reconciliation, no support ticket asking "where's my reward?"
Customers who travel don't want to pay a full monthly rate for a line they aren't using — but they don't want it cut off either. The portal lets them schedule a reduced-rate holiday period themselves: pick a plan, choose a start date and a duration of one, two, or four weeks, and see the price before confirming. The line stays alive at a lower cost for exactly the window they're away, with no phone call and no manual billing adjustment from staff.
Leaving should be as self-service as joining — and just as hard to do by accident. A customer can request termination from the portal; the system applies the correct notice period automatically and then confirms intent through a second channel before anything is final. Accounts with two-factor enabled confirm with their 2FA code; everyone else gets a time-limited secure link emailed to the address on file, so a single mis-tap never cancels a service. Changed their mind? Cancelling the termination is the same one-step, self-service flow, and the customer keeps their service uninterrupted.
The cheapest way to fix a churn problem is to hear about it before the customer leaves. A short satisfaction survey reaches new customers inside the portal during their first weeks of service, asks how the install and early experience went, and routes the answers straight back to the team. It takes the customer two to three minutes, opens only inside its intended window, and turns a quiet first impression into something the operator can actually act on.
Customers update their own portal email address themselves — the change only takes effect after they confirm via a link sent to the new address, with safeguards that prevent guessing whether an address is already in use.
Customers manage their own home network from the portal. They rename their WiFi network and change its password — pushed straight to the fibre device, not read down the phone — register or swap the router MAC addresses on their line, and run a one-tap connection test that reports whether the router is online and which device is responding. Every change is logged to the account and confirmed by email, so the helpdesk never has to talk a customer through router settings.
Customers pull up any past invoice, credit note, or proforma in the portal and download it as a PDF themselves, with every document scoped to their own account. No more emailing support to re-send a copy of a bill.
The service view surfaces a plain-language fibre signal indicator — good, weak, or very weak — drawn from live network diagnostics, so a customer can see whether their line is healthy before they ever open a ticket. It turns the operator's own monitoring data into something the customer can read for themselves, without exposing the raw engineering numbers behind it.
The situation. A residential customer's internet stops working at 9:47 PM on a Saturday. Support office is closed; the after-hours phone tree exists but the customer would rather not navigate it.
What the customer does. Logs into the portal on their phone. Portal Badger automatically runs the diagnostic: ONU signal is nominal, OLT port is up, the issue is most likely a router-level reboot. Portal offers "Reboot ONU" with one tap. Customer taps. Within 30 seconds, the ONU reboots and reconnects. Customer's phone shows a green checkmark.
The result. Service restored at 9:48 PM. No phone call placed. No ticket opened. The customer leaves the interaction thinking "that worked." The Monday morning support team sees zero tickets from this customer; they would never have known about the issue at all if it had not appeared on the Portal Badger usage report.