Logo Dodo

Logo Dodo is a local-first logo workbench for moving from a short brand brief to generated concept boards, structured candidate feedback, and increasingly focused visual directions. The workflow is built around practical art direction: generate a broad board, mark the strongest and weakest candidates, then use that feedback to steer the next round.

• Live site: logododo.com
• Example walkthrough: Rise and Riot

Workflow

Logo Dodo turns a brand brief into a sequence of reviewable boards:

• Creates 5x5, 4x4, 3x3, and 2x2 logo concept boards.
• Detects candidate bounds so individual marks can be reviewed apart from the full board.
• Captures likes, loves, rejections, tags, notes, and additional reviewer direction.
• Synthesizes feedback into the next creative territory for iterative generation.
• Exports a Markdown creative summary with the brief, board history, selected notes, prompts, and run metadata.

How it works

The app uses SolidStart with server-side OpenAI calls:

• Text models plan visual territories and refinement directions.
• Image generation produces PNG board assets saved into the local app workspace.
• Candidate previews are extracted as a review convenience layer.
• Project history and feedback are persisted locally so each board has durable context.

Highlights

• Structured brand briefs with include and avoid guidance
• Iterative board generation from user feedback
• Candidate-level review and tagging
• Local project history with Markdown handoff export
• Docker Compose setup with persistent data and generated asset volumes

Overview

Why it exists

Evolution and milestones

What changed

Lessons

What I would do differently

Next