Hobbyboard turns folders of photos, sketches, and screenshots into a searchable archive with semantic search, rich tags, and beautiful boards — all on your own machine.
Joinery, finish notes, and reference photos in one place.
Semantic search across captions, OCR, and tags.
Your media stays on your machine. You choose the AI backend.
The same visual language as the app: dark, warm, and focused.
Create curated boards while keeping ad‑hoc tags for quick filtering.
Search by intent, not file name. Find “arched legs, brass hardware” instantly.
SQLite + Qdrant, with embeddings generated locally via fastembed-rs.
Automatically extract titles, tags, and visible text for richer discovery.
Point Hobbyboard at a folder of inspiration and let it do the rest.
Creates the database, models folder, and vector collection.
Scans media, generates thumbnails, runs AI analysis, and embeds.
Launches the web UI with hybrid semantic + text search.
Full docs live at hobbyboard.aravindh.net/docs.