What are the best ways to share SQL/Python query results with external users?

0
18
Asked By CuriousCoder92 On

I'm currently working with SQL and Python to query Oracle and Impala databases, then send those results to Google Sheets, which I use with Looker Studio for dashboards. While this setup works, it feels a bit clunky, especially when I want external users to have the ability to filter data themselves, like by Client ID. I'm looking for alternatives that would allow me to publish tables and charts more straightforwardly, while still letting users safely run parameterized queries. Would transitioning to something like Streamlit or FastAPI with JavaScript be a good move? I'm eager to hear what solutions others have found effective!

6 Answers

Answered By ExcelExpert5 On

Have you considered using Power Query in Excel or even Power BI? This solution is designed for just this kind of task, letting you save the data embedded and giving users the ability to manipulate it directly.

Answered By OpenSourceFanatic On

Instead of Power BI, you might find the open-source framework Evidence useful. It lets you write SQL queries against DuckDB and pre-loads them for easy reuse. You can create dashboards using a Markdown/HTML-like syntax, plus you get version control through Git. The end result is a static HTML dashboard that’s pretty clean and efficient.

Answered By DashboardsRUs On

Have you tried using Grafana? It's another excellent option for building interactive dashboards if you want to consider alternatives beyond just tables and charts.

Answered By TechSavvySam On

I set up a backend API that sends data from SQL or files to my frontend. To allow users to filter and run complex SQL queries, I utilize DuckDB with WebAssembly and some JavaScript. It runs the SQL engine directly in the browser, so I don’t have to worry about backend query sandboxing. It's been a game changer for user interaction.

Answered By DataWizard007 On

If your goal is just to let users view datasets and create dashboards/visualizations easily, you might want to check out Metabase. It has data sandboxing features, which means you wouldn't need to start from scratch to build an application. It's a solid choice for giving users a simple interface for their queries.

Answered By BigDataGuru On

The best approach really depends on who your audience is. For my work in Big Data, I often export tables to parquet files, which researchers can then load into Python for further analysis. It's a straightforward way to handle large datasets.

Related Questions

LEAVE A REPLY

Please enter your comment!
Please enter your name here

This site uses Akismet to reduce spam. Learn how your comment data is processed.