The network for creativity
Join 1.25M professional creatives like you
Connect with clients, get discovered, and run your business 100% commission-free
Creatives on Contra have earned over $150M and we are just getting started
Back to feedPost
One of our clients needed 70 ad video variants in a week, so we built an AI agent to do it in seven hours.
The build itself was a project. It required:
- Assembling a Claude agent connected to Remotion
- Building a knowledge base of the client's brand: brand guide, movement language, hook creation guidelines, past deliverables, briefing structure
- Teaching the agent to read the brief the way our team would
- Composing hook variants adapted for each aspect ratio (1x1, 9x16, 16x9)
- Assembling endcards and outputting files named and organized for QA
The first batch still needed heavy manual editing. The second, the agent assembled alone.
The third came out fully stitched (hook, body, CTA, disclaimer) needing only a review pass before shipping.
Each iteration meant testing what the agent got wrong, adjusting the skills, refining the knowledge base, and rerunning until the output matched the standard our team would ship.
When we sent the batch, the client feedback was: "The revised videos look great. We'll start using them immediately."
To show what this workflow can do, the video in this post was produced by this exact system.
If your business needs to ship ads at this volume, let's talk 👉 apta.agency/contact
70 ad variants in seven hours is wild. remotion?
Yes! A framework of Claude working directly with Remotion.
Assembling a Claude agent connected to Remotion to hit a seven-hour turnaround is genuinely solid infrastructure work. What almost broke when you scaled from a handful of variants up to 70?
The network for creativity
Join 1.25M professional creatives like you
Connect with clients, get discovered, and run your business 100% commission-free
Creatives on Contra have earned over $150M and we are just getting started
Challenges
View all
Related posts
I recreated an ElevenLabs ad shot-for-shot using AI. Same script structure, same camera moves, same phone UI interaction, different product, different character, built from scratch.
Here's the part people usually skip when they talk about "using AI" for video:
I didn't type one prompt and get a finished ad. I found a look I wanted on Pinterest, ran it through ChatGPT to turn it into a structured character description, then used Nano Banana 2 to build a full character sheet, so she stayed consistent across every single shot. That consistency problem is the actual hard part of AI video; anyone can generate one great frame.
From there, each scene was generated in Seedance 2.5 through Magnific. The phone UI you see her interacting with? I designed that myself, frame by frame. Motion graphics and text overlays went through After Effects. Sound design and SFX in CapCut. The voice was ElevenLabs v3, also run through Magnific.
Six tools, one continuous pipeline, one consistent character, one final cut.
This is the actual value I bring to clients, not "AI made a video," but knowing how to direct a full production through AI: casting, consistency, UI/UX detail, motion, sound, all held together with intent. That's the difference between content that looks AI-generated and content that just looks good.
If you're a brand wondering what's actually possible with AI production right now, this is a good example of the ceiling. Happy to walk anyone through how it's done.
Crafted a few surreal frames with GPT 2.5 Sunburst, then brought them to life using Seedance's multiframes feature.
SaaS Explainer Video — AI-Native Software Engineering.
A cinematic SaaS explainer created to communicate a simple business message: AI can build an app, but building software that works reliably for a real business takes much more.
The video combines an AI presenter, product interface visuals, technical environments, and cinematic storytelling to explain the difference between a quick AI prototype and production-ready software.
The goal was to make a technical SaaS product feel clear, modern, credible, and easy to understand while keeping the story engaging from beginning to end.
Tools & techniques: AI video generation, AI visuals, product visualization, cinematic storytelling, motion design
AI ArtAI Content CreationGoogle FlowGoogle LabsRunwayaivideoproductionaiartistaivideoproducerAI Video Production
the saas shot is clean.