Part Time
$2500/m
TBD
Jan 11, 2026
PLEASE APPLY FOR THIS ROLE HERE:
What we do:
We install AI creative systems into e-commerce brands and agencies. We're are not a SaaS company or a course, this is custom-built infrastructure that lives inside our clients' businesses and produces them ads at scale. E.g.: a client comes to us spending $100K/month on ads but bottlenecked at 20 creatives per week - we build them a system that produces 200 ads, with the same team and budget.
We build creative machines that make ads, then hand over the keys.
The role:
You'll be the person building these systems. Not configuring, not implementing someone else's designs - building them all from scratch. Taking a client's requirements and turning that into a production workflow that runs reliably, scales, and doesn't break.
You'll work directly with me - we're a small team with no corporate layers, and you'll have direct access to clients when needed.
What you'll actually build:
Real examples of systems we're building right now:
Competitor Intelligence Pipeline
Input: Client's website URL
System scrapes site for brand info, searches for competitors, pulls their ads from
Scores ads by effectiveness (days running × variations)
Downloads winning video ads, extracts scripts via Gemini
Claude analyzes scripts for psychological patterns and hooks
Outputs: Spreadsheet tracker + pattern analysis + rewritten scripts for client's brand
AI Ad Generation Engine
Input: Product URL + brand guidelines
System generates ad concepts, writes scripts, creates UGC avatar images
Animates avatars speaking scripts (Veo 3.1)
Generates B-roll scenes, animates them (Kling)
Outputs: Folder of production-ready video ads with proper naming/metadata
Creative Variation System
Input: One winning ad
System analyzes what makes it work, generates 20 variations
Different hooks, different CTAs, different visual treatments
Outputs: 20 ready-to-test ad variations
Tech stack:
Primary:
n8n - This is the core. You must be EXCEPTIONAL at using n8n.
APIs - You'll integrate 15+ services and be able to debug auth issues in your sleep.
AI APIs - Claude, Gemini, OpenAI. Prompt engineering, structured outputs, handling rate limits and failures gracefully.
Secondary:
Gumloop - As an alternative to n8n for certain builds
Python - When n8n can't do what you need (it happens)
Google Sheets/Drive - Client-facing outputs
Airtable - Some client systems live here
AI Creative Tools (you'll integrate these, not use them manually):
Arcads, HeyGen, NanoBanana, ElevenLabs (voice/avatar)
Kling, Runway, Veo 3.1 (video generation)
Midjourney, DALL-E, Flux (image generation)
Scrape Creators, Apify (data extraction)
You don't need to know all of these. You need to be able to read API docs and figure them out fast.
What we're looking for:
Non-negotiable:
You've built MULTIPLE production workflows in n8n from scratch that other people use.
You understand APIs deeply. You've authenticated against OAuth2 APIs. You've parsed nested JSON. You've handled pagination. You've debugged why something returns 401 when it should return 200.
You're extremely fast.
You communicate clearly. When something's blocked, you say so immediately. When something will take longer than expected, you say so before the deadline, not after.
You're resourceful. You Google, you read docs, you find solutions. You don't wait to be told how to do things.
Strong signals:
You've built automations specifically for content or creative generation
You've worked with AI APIs and understand prompt engineering
You've touched some of the creative tools (Arcads, HeyGen, etc.) even if just experimenting
You have a portfolio or GitHub that shows real work
You've worked in fast-moving startups or as a freelancer (you know how to operate without structure)
Red flags (don't apply if this is you):
You need detailed specs before you can start building
You're uncomfortable with ambiguity
You think "that's not my job" about anything
You disappear when you hit a problem instead of communicating
You want a 9-5 with clear boundaries
How we work:
Slack for communication. Expect responses within a few hours during working hours. Same expected from you.
Loom for walkthroughs. You'll record your builds. We'll record feedback.
Weekly sync call if needed. Otherwise async.
You'll have direct access to client calls when context is needed.
When something's urgent, it's urgent. When it's not, take your time and do it right.
Compensation:
This is a contract/freelance based role to start
Hourly rate based on experience
Potential for retainer or full-time if you're exceptional and want it
To apply:
Fill out this application form:
If you don't record the Loom video, you will be automatically rejected. We need to see how you think and communicate. If your Loom is good, we'll send you a trial task.
Don't apply if:
You're looking for a stable corporate job with clear career progression
You're not comfortable moving fast and figuring things out
You can't show us real work you've built
You need more than 48 hours to turn around a trial task
Do apply if:
You read this and got excited, not intimidated
You've built things you're proud of and want to show them off
You want to work on genuinely interesting problems, not boring automation
You're looking for a team that will push you and let you push back