← All Projects

Figma Weave
AI Workflow

Mosaic
solo build AI workflow
AI Systems Brand
The Problem

Producing a single social asset at Mosaic meant sourcing imagery, drafting copy, looping in collaborators, and manually applying brand rules — a process that averaged 2 days per post for what should be routine output. With a recurring social calendar and a lean team, that bottleneck wasn't sustainable.

2d Previous turnaround per asset
4h Turnaround after workflow
1 Person. Solo build & maintenance
The Solution

Built a Figma Weave AI pipeline that takes a social calendar brief as input and outputs production-ready, on-brand assets — no manual image sourcing, no copy pass, no back-and-forth. The system handles categorization, brand application, and layout generation end-to-end.

How It Works

The brief gets ingested and converted into structured JSON — headline, body copy, and CTA — then automatically categorized by post type: community investment, operations, recruiting, sustainability, and more. Each category maps to a pre-defined background treatment so the system knows which visual direction to apply without any manual input.

JSON structure fed into the Figma Weave workflow
Brand Guardrails

The pipeline was trained on Mosaic's brand guidelines, approved typography, color system, and a curated library of reference assets. Every output stays on-brand without a manual review pass — the system enforces the rules so I don't have to.

Reference library of approved brand assets fed into the workflow
Output Examples

Generated assets across post categories — each one brand-compliant and production-ready, no manual intervention.

AI-generated social asset — community investment post AI-generated social asset — operations post
Result

Asset turnaround dropped from 2 days to 4 hours — built and maintained solo. Time recovered went back into higher-priority design work instead of repetitive production cycles.