Getting Started
Get your brand indexed and generating on-brand AI content in under 10 minutes.
Step 1: Create a brand
After signing in, click New Brand from the dashboard. Give it a name (internal only — this is how you'll identify it in uvoo).
Step 2: Connect your assets
Choose your ingestion method:
- Instagram handle — enter your public username (e.g.
@yourbrand). uvoo scrapes your feed and downloads full-resolution images. - Website URL — provide your homepage or a specific gallery URL. We crawl linked images.
- File upload — drag and drop a ZIP archive. Supported formats: JPG, PNG, WebP.
Ingestion typically completes in 3–8 minutes depending on asset count. You'll see a live progress bar.
Step 3: Wait for analysis
Once assets are ingested, uvoo automatically runs the full analysis pipeline — color extraction (CIELAB), visual embeddings (SigLIP 2), semantic tagging (RAM++), and composition analysis. For 200 assets, this takes about 5 minutes.
You'll receive a notification when analysis is complete. Your brand knowledge graph is now ready.
Step 4: Try Reference → Prompt
Go to the Reference Guide tab. Upload any image — from your own library or an external reference — and uvoo generates brand-tuned prompts for Midjourney, DALL-E 3, Stable Diffusion, and Ideogram.
Each prompt includes your brand's actual palette values, lighting description, composition parameters, and mood tags — all derived from your analyzed assets.
Step 5: Ask your Brand Assistant
Go to Brand Assistant and try a few natural language queries:
- "What colors dominate our product photography?"
- "What lighting style do we use most often?"
- "Which assets are visual outliers for our brand?"
The assistant answers from your actual asset data, not from generic brand knowledge.
What's next
- Read Brand Signals to understand all 58+ extracted fields
- Read Prompt Formats to see exactly how signals translate into model-specific prompts
- Read API Reference if you want to integrate uvoo into your existing workflow