SouravRoy-ETL

SouravRoy-ETL / duckle

Public

Local-first ETL/ELT studio: a drag-and-drop visual pipeline designer that compiles to SQL and runs on DuckDB. Tiny desktop app, no servers, git-friendly workspaces.

data-engineering data-integration data-pipeline data-quality desktop-app
16
1
100% credibility
Found May 25, 2026 at 16 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Rust
AI Summary

Duckle is a free, open-source desktop application that lets you build and run data pipelines visually on your own computer. You drag sources like CSV files or databases onto a canvas, connect them through transforms like filters or joins, and wire them to outputs. An AI assistant can generate pipelines from plain-English descriptions. Everything runs locally through DuckDB with no cloud dependency, and your work saves as plain files you can version with Git.

How It Works

1
📥 Download and install Duckle

You download a small 30 MB app for your computer. On first launch, it asks to install the data engine (about 30 seconds) and optionally the AI assistant (~1.1 GB).

2
📁 Choose where your projects live

You pick any folder on your computer as your workspace. Everything you build stays there as plain files you can open, share, or back up with Git.

3
🎨 Build your first pipeline visually

You drag a CSV source onto the canvas, drop a Filter in the middle, and wire it to a Parquet output. Click any node to see the generated SQL or preview your data.

4
✨ Ask the AI to build it for you

Click the sparkles icon and type 'read orders.csv, filter where status is paid, save to paid.parquet.' The AI assistant draws the pipeline on your canvas in one click.

5
▶️ Run and watch your data flow

Press Run. Each step lights up as it executes. You see row counts, timing, and any errors. Stop mid-run anytime. Your results land exactly where you specified.

6
🔄 Schedule it to run automatically

Attach a schedule to run nightly, every 15 minutes, or whenever a new file arrives. Duckle wakes up and runs your pipeline even while you sleep.

🎉 Your data work is done locally

No cloud accounts, no API keys, no monthly fees. Your pipelines, your data, your computer. Everything is open source and yours to keep.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 16 to 16 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is duckle?

Duckle is a desktop ETL/ELT studio that lets you build data pipelines visually. You drag components onto a canvas, wire them together, and run. The tool compiles your graph to SQL and executes it through DuckDB, so you get native speed without a server. Built in Rust with a React frontend wrapped in Tauri, the whole app ships as a single ~30MB binary. It includes an AI assistant called Duckie that generates pipeline JSON from English descriptions, and workspaces are plain files you can commit to Git.

Why is it gaining traction?

The local-first promise is real here. No cloud, no API keys for the AI assistant, no telemetry. Duckie runs entirely on your CPU using a small Qwen model. The connector count is impressive: 290 sources covering databases, cloud warehouses, SaaS APIs, streaming systems, and vector stores. Everything is self-contained and auditable. The fact that pipelines are plain JSON files means you get real version control, diffs, and branch workflows without a proprietary lock-in.

Who should use this?

Data engineers who want a visual way to prototype ETL logic without spinning up Airflow or dbt. Analysts who know SQL but prefer clicking to coding. Small teams that need to move data between PostgreSQL, S3, Snowflake, or pgvector without infrastructure overhead. Anyone doing RAG ingestion pipelines who wants to chunk, embed, deduplicate, and land data in a vector store through one canvas.

Verdict

Duckle is a credible early-stage project with a well-thought-out feature set and clean architecture. However, with only 16 stars and a 1.0% credibility score, the community is essentially nonexistent. The documentation is thorough and the Rust codebase looks solid, but you are an early adopter. Watch the releases closely and test thoroughly before betting production workflows on it.

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.