Department: Customer Support

Location: On-Site — H-12, NSTP, Islamabad

Reports to: Founder / Head of CX

Type: Full-time

About ImagineArt

We're redefining how the world creates and designs.

ImagineArt is one of the fastest-growing GenAI companies in the world. We've scaled faster than most funded startups, with zero outside funding — 100M+ social impressions, and our own image generation model ranked #3 globally for photo realism.

No funding. No shortcuts. Just a sharp, driven team building one of the strongest GenAI products in the world, and we're just getting started.

About the Role

This is not a ticket queue job. We're hiring the person who owns customer support at ImagineArt — the standards, the team, the tooling, and the numbers.

Millions of creators use our image and video tools across web and mobile, in dozens of languages, on every kind of device. Today that support is held together by hustle. You're the person who turns it into a function: a team that scales, a technical escalation path that engineering trusts, and an AI-first operation where the boring 60% of tickets never reach a human.

You'll be hands-on for the hardest cases — failed generations, payment failures, account and credit disputes, platform incidents — and you'll be the person we call when something breaks at 2am. But your real output is leverage: fewer tickets, faster resolution, and a team that gets better every month.

What You'll Own

The team

Hire, onboard, coach and performance-manage support specialists as we scale

Build the L1 → L2 → engineering escalation ladder, and decide what sits at each rung

Run QA on replies and tickets — define what "good" looks like, then enforce it

Own scheduling and capacity so coverage holds across timezones without burning people out

Write the onboarding so a new hire is independently handling tickets in two weeks, not two months

The technical layer

Personally own L2 escalations: reproduce bugs, pull logs and screenshots, write bug reports engineering can act on without a follow-up

Run incident comms during outages and widespread failures — status updates, user messaging, post-mortems

Partner with engineering and product to get fixes shipped, and close the loop back to affected users

Handle the ugly edge cases: refunds, chargebacks, credit disputes, App Store and Google Play billing failures, account recovery

The AI and automation layer

Build the deflection machine: macros, chatbot flows, auto-tagging, smart routing, saved replies

Own the AI support bot end to end — training it, correcting it, keeping it current as we ship features weekly

Maintain the help center, FAQs and internal playbooks as living documentation, not a graveyard

Ruthlessly hunt repeat tickets and kill them at the source — better UX copy, a help article, a product fix

The numbers

Own CSAT, first response time, resolution time, deflection rate, reopen rate, and cost-to-serve per ticket

Build the reporting that makes those numbers visible weekly, and act on them

Be the voice of the user into product: what breaks, what confuses, what people are actually trying to do and can't

Must Have

4+ years in customer support, with at least 1–2 years managing or lead-level ownership of a support function — team, metrics, or both

Excellent written and spoken English. You'll write user-facing copy and you'll be judged on it

Genuine L2 ability — comfortable reproducing bugs, reading logs, distinguishing a user error from a product bug from a payment gateway failure

Strong technical fluency: web apps, mobile apps, browsers, cache and session issues, basic troubleshooting across devices

Hands-on AI tool use in daily work — ChatGPT, Claude, or equivalent. You should already have opinions about where they help and where they quietly make support worse

Support platform experience — Intercom, Zendesk, Freshdesk or similar, at admin level, not just agent level

Calm under pressure, patient, and solution-focused. Support escalations are emotional; you can't be

Good to Have

Automation tooling — Zapier, Make, or n8n

Bug tracking — Jira or Linear