# I wrote a Parquet viewer in Rust to avoid running SQL for the PM

Every data team knows the drill: a PM needs to “just take a quick look” at some Parquet data. That usually means asking an engineer to write SQL or spin up a tool to pull a few rows. It’s a small ask, but one that happens often enough to slow everyone down.

That’s why I built [**Tablr**](https://github.com/lzm0/tablr) — an extremely simple desktop app for browsing Parquet files quickly and without any coding.

It’s a native desktop app written in Rust, powered by the blazing-fast [Polars](https://github.com/pola-rs/polars) library and the [`egui`](https://github.com/emilk/egui) UI framework, and it runs on macOS, Windows and Linux.

### Key Features

Here’s what you get out of the box:

* 🗂️ **Multi-file support** – Load single files or partitioned datasets with ease.
    
* ⚡ **Infinite scrolling** – Efficiently scroll through massive datasets, thanks to Polars’ lazy execution.
    
* 🔍 **Basic filtering and sorting** – Click on a column to sort, or filter rows using simple operations like “equals” or “contains”.
    
* 💻 **Native performance** – Built with Rust for speed and low memory usage.
    
* 🌐 **Cross-platform** – Works on Linux, macOS, and Windows.
    

### How to Try it

There aren’t prebuilt binaries yet, but if you’ve got Rust installed, it’s easy to get up and running:

```bash
git clone https://github.com/lzm0/tablr
cd tablr
cargo run --release
```

One of my favourite features is that you can open Parquet files straight from your terminal — no GUI navigation required. Just point it at a `.parquet` file:

```bash
tablr path/to/your_file.parquet
```

Or, if you’ve got a folder of partitioned files:

```bash
tablr path/to/your_dataset/*.parquet
```

### Get Involved

For ideas, bug reports and pull requests, jump in here:

👉 [https://github.com/lzm0/tablr](https://github.com/lzm0/tablr)