-
-
TYPE OF WORK
Full Time
-
-
SALARY
$1000
-
-
HOURS PER WEEK
40
-
-
DATE UPDATED
May 5, 2026
We run LinkedIn marketing for 45+ B2B clients.
Every piece of content we publish ships with a custom graphic. Carousels, infographics, personal-brand posts, company-page graphics. All on-brand. All made fresh.
The bottleneck isn't ideas. It's design throughput.
I need someone who owns this pipeline end to end.
This isn't a designer job. It's a process operator job. You know what good looks like, you know how to fix what's broken, and you know how to ship at volume without quality dropping.
If you know Claude, Canva, and Nano Banana, and you actually use them in your workflow rather than nodding at them in interviews, keep reading.
What you'll be doing
LinkedIn carousels at volume. Four to eight slides, multiple per client, every month, 45+ clients
Infographics that earn the save. Clean, dense, on-brand
Personal-page and company-page graphics matched exactly to each client's brand kit
Running AI image tools (Nano Banana, GPT Image 2) inside an actual workflow, not as a novelty
Maintaining 45+ brand kits in Canva without ever mixing them up
Who you are
You build reusable carousel skeletons and components so you're not solving the same problem twice.
You can turn a six-slide carousel around in an hour, not days.
You're already on top of how Claude, ChatGPT and similar tools fit into a designer's stack. You use them to plan layouts, write microcopy, draft variations, and speed up everything that doesn't need a human hand.
Nano Banana is already in your workflow. Not something you'll "look into."
You take review and feedback cleanly. Your work gets checked, consistently.
Canva is a tool you respect, not one you look down on.
Must-haves
You know how to prompt. im not here to give you free training
Portfolio showing LinkedIn carousels at volume. Not one or two hero pieces
5+ years Canva. Brand kits, templates, team workflows
Hands-on AI image generation in a real production workflow
Strong infographic design craft
Managing multiple brand identities concurrently
Full-time availability (40 hrs/week effective)
Nice-to-haves
Figma. Light motion or animation. Prior agency or batch-production experience.
One last thing
This role only works if you're detail-obsessed. 45+ brand kits means 45+ ways to get it wrong.
I'm looking for someone who treats consistency as the job, not a chore.
If that sounds like you, apply.
Screening Questions
Q1 — Volume proof
In your last 30 days, how many LinkedIn carousels did you ship, across how many different brands? Drop links to three examples, each from a different client. We're looking for proof of throughput, not hero pieces.
Q2 — Brand kit system
You're juggling 10+ brand kits and a client messages saying their accent colour got swapped in last week's batch. Walk us through your actual process for: (a) verifying it happened, (b) fixing it across what's already published, (c) making sure it doesn't happen again.
Q3 — AI workflow proof
Pick one LinkedIn carousel you made in the last month where you used AI tools (Nano Banana, Claude, GPT Image 2, or similar). Walk us through specifically: what tool, what prompt, what output, what you kept, what you edited, how long the whole thing took. Specificity matters. "I use ChatGPT for ideas" is not an answer.
Why these three:
Q1 filters out portfolio-padders. Most candidates will pad with two beautiful hero pieces and hide the volume question. The right candidate has the numbers ready and the links queued.
Q2 filters out designers pretending to be operators. Anyone without an actual system will give a vague answer about "double-checking." The right candidate has a checklist, a naming convention, or a workflow they'll describe in concrete steps.
Q3 filters out AI-tourists. Most people will write something generic about "speeding up workflow." The right candidate gives you a screen-by-screen breakdown with real prompts and timestamps.
Bonus filter worth considering:
If Indeed lets you ask a fourth, this one cuts hard:
This isn't a creative-direction role. The frameworks, copy, and visual systems are already defined. Your job is to execute them at volume with consistency. Why does that excite you?
