Case Study
How I built an internal
AI production process
for OpenSea.
My Role
Deliverables
Overview
OpenSea had recently made a company-wide pivot toward AI-driven content production, but there were no systems or structure in place to support it.
To fill that gap, I began developing an internal AI production process on my own time. The goal was to create clarity, consistency, and efficiency where none existed, and to lay the foundation for a scalable workflow the entire team could use.
OpenSea wanted to build an AI content team,
but first it needed structure.
When I joined, there was no consolidated list of cleared-to-use IP.
OpenSea wanted to create content that promoted the marketplace, but there was no official guidance on which NFT or token IP we were legally allowed to use. Instead, the direction was to reference the recent brand video and pull IP from there.
For obvious reasons, this created major inefficiencies in the production process.
From video to video, the AI character consistency was nonexistent.
I believed it was our responsibility, as stewards of the marketplace, to maintain the integrity of the communities, IPs, and digital art we represent.
Brand assets were scattered across departments, Google Drives, and Figma files, making production and onboarding slow and disorganized.
I saw an opportunity to bring
real value
to the organization.
I created an OpenSeapedia,
the internal hub for everything related to AI content.
Bringing all resources together under one website ensured the entire team had access to the most up-to-date assets, tools, and internal guidance.
This was a self-initiated side project that I built primarily over nights and weekends.
Visit Demo Site
A Hard-Working 32-Page Internal Resource
IP Resources
To solve the issues around IP usage and character consistency, I developed comprehensive IP profiles with curated character sets for training AI content.
3D_3/4
3D_Front
2D_Front
Pixel_3/4
Pixel_Front
The Process
I conducted deep research on each
Web3 brand to determine how they
stylistically depict their base characters.
Using a combination of AI and
traditional illustration methods, I
created precise character perspectives.
Lastly, I developed guidelines for how all the
base characters coexisted within the same ecosystem.
System In Practice
The base characters could be used to generate alternate character sheets that
remained on-brand and stylistically consistent with the original IP.
remained on-brand and stylistically consistent with the original IP.
This process allowed for greater creative freedom while ensuring video-to-video
character consistency. It also made production faster and more efficient.
character consistency. It also made production faster and more efficient.
Watch Demo Video
A breakdown of the internal AI production system.The Results
By centralizing assets into a living, always-up-to-date resource, defining clear IP profiles, standardizing character systems, and building a modular AI workflow, I transformed a fragmented process into a streamlined internal production system.
The team gained structure, consistency, and speed, turning a chaotic workflow into a simple, repeatable system that enabled OpenSea to scale its AI content efforts with confidence.