Support without escalation

Most escalations from support are not bugs. They are questions about why the system did something: a fee, a hold, a refund that went out twice. The answer sits with one developer, and the customer waits a day for it.

The usual escalation

1. Customer: “I’m on the payment holiday you approved in September. Why was I charged a ₹590 late fee?”
2. Support: The fee looks deliberate. Escalates to engineering.
3. Engineering, next day: A developer reads the fee rule, finds it checks days past due and nothing else, and replies that the fee is “as designed”.
4. Support defends a fee the loan terms forbid. The customer complains again.

With AppThentic

Developers explain each flow to om once. After that, support asks om.

1. Support asks om from the ticket, in plain words.
2. om answers with the rule and its owner, and what to tell the customer.
3. Support resolves the ticket in two minutes. No developer is pinged.

When om doesn’t know

om says so, asks the rule’s owner, and comes back to the ticket with their answer. The next agent who gets the same question gets that answer straight away.

Answers from the people who know

Anyone can add what they know. Priya S., Payments lead, adds a note: a customer refunded twice is almost always a gateway retry, so check the idempotency key on both refunds first. Three weeks later she is on leave, and om answers Tara in Support the way Priya would, credited to her note. See Organization Memory.

What support sees

Admins assign who sees what. Support gets the customer impact and the next step, not the stack trace, and never anything outside its scope. See Access and audit trail.

What it saves

Escalations times developer minutes, handed back. A ticket that took a day to escalate closes in two minutes. The questions support asks most go to their owners first, so the next ticket is answered too.

Who uses it

Heads of support and operations, and the support agents who answer customers about money, holds, fees and refunds.