🦬 zubr

a modern, powerful desktop gui for postgresql

---

the for now not so ultimate desktop client for postgresql, supercharged with vector visualization for ai/ml.

perfect for developers, data scientists, and anyone working with pg_vector and modern data applications.

a note on technology: i’m not a huge rust fanboy, but i am a fan of my ram. zubr is built with tauri instead of electron to keep it lightweight and fast.

✨ features

• ✅ modern & fast ui: a clean and responsive user interface built with react and tailwindcss.
• ✅ direct postgresql connection: connect securely and directly to your local or remote postgresql databases.
• ✅ schema & table navigation: easily browse schemas and tables in a dedicated sidebar.
• ✅ vector embedding visualization: a fancy feature to inspect pg_vector embeddings with a heatmap, stats, and raw values.
• ✅ cross-platform: built with tauri to run natively on macos, windows, and linux.

🏗️ architecture

graph TD
A["user<br/>(react frontend)"] --> B["zubr core<br/>(tauri/rust)"]
B --sql commands--> C["postgresql<br/>database"]
C --query results--> B
B --ui state--> A

style A fill:#b7bdf8,color:#24273a
style B fill:#f5bde6,color:#24273a
style C fill:#8bd5ca,color:#24273a

Loading

how it works

• frontend: a modern react application provides the user interface for all database interactions.
• core: the tauri backend, written in rust, handles the native window, os integrations, and manages the direct connection to the postgresql database.
• postgresql: zubr communicates with any standard postgresql database, with special features for the pg_vector extension.

🗺️ roadmap

✅ what works now

• ✅ database connection: establish a connection to a single postgresql database.
• ✅ schema/table browsing: view a list of schemas and the tables within them.
• ✅ data viewing: display table data in a clean, readable grid.
• ✅ vector visualization: inspect vector embeddings from your table.

🔄 working on next

• 🔄 connection manager: save and manage multiple database connections.
• 🔄 in-cell data editing: edit data directly within the table grid.
• 🔄 advanced filtering: a dedicated ui to build complex where clauses without writing sql.

📋 planned improvements

• 📋 sql query editor: a full-featured sql editor with syntax highlighting and autocompletion.
• 📋 tabbed interface: open multiple tables and queries in separate tabs.
• 📋 vector similarity search: a ui to perform k-nn similarity searches on your vector columns.
• and many more ...

legend

• ✅ completed - ready to use
• 🔄 in progress - actively being developed
• 📋 planned - scheduled for development

🤝 contributing

this project is not yet ready for contributions, but please check back soon! when it is, we’ll be looking for help with:

1. bug fixes and reporting.
2. implementing new features from the roadmap.
3. improving documentation.

but if you want, you can already start testing it and making issues

📄 license

this project is licensed under the mit license.