There are plenty of CSV/Parquet viewers out there, but I kept running into small frustrations that added up. After bouncing between Excel, Numbers, VS Code, and a bunch of web tools, I realized none of them quite fit how I actually work. So I decided to build one myself using DuckDB-Wasm + Svelte.
I built this mainly for my own workflow. A few things I really wanted:
1. SQL-first exploration + export. When I'm working with multiple CSVs or messy schemas, clicking around only gets me so far. Being able to write SQL (or programmatic queries) is the fastest way for me to understand the data, and I want to export query results easily.
2. High performance on large files. I love the Rainbow CSV plugin in VS Code, but once files hit GB scale, my laptop starts to struggle. Apple Numbers can take minutes just to open a file, and it doesn't support SQL at all. I wanted something that stays responsive even for large datasets.
3. A proper GUI for quick inspection. Sometimes I don't want to write SQL, I just want to sort, filter, and search with a few clicks.
4. Infinite scrolling. This is surprisingly rare. Most web CSV tools paginate results, which constantly breaks my flow. I really wanted to just scroll through the data naturally.
This is still early and very much a work in progress, but it already covers the core things I personally needed. I'd love to hear any feedback, especially if you work with large CSV/Parquet files or have strong opinions on data tooling
Thanks for checking it out!