Analyse data without uploading it
Most analysis tools ask you to hand over the file first. Malytics does the work in the browser tab you already have open, which changes what you are allowed to use it on.
If your data is a customer list, a payroll export, a patient extract or anything under an NDA, the blocker is rarely the analysis. It is that uploading it to somebody else's server is a decision you are not allowed to make on your own.
What "it stays in the browser" actually means
Your file is read by the page, parsed in memory, and stored in your browser's own IndexedDB — the same place a web app keeps your drafts. Queries, transforms, charts and statistics all run as JavaScript in that tab. There is no upload step because there is no server-side copy to upload to.
The practical consequences:
- Close the tab and clear site data and the dataset is gone. Nobody has to be asked to delete it.
- It works offline. Once the page has loaded, analysis does not need the network.
- Your file size is limited by your machine, not by an upload quota.
- No data processing agreement is needed for the analysis itself, because no third party processes it.
The one exception, stated plainly
Ask the built-in AI a question and it sends three sample rows and your column statistics to the model — not the file. It sends nothing at all until you ask it something.
If that is still too much, bring your own API key, or do not use the AI features. Everything else works without them.
We are specific about this because "private" is a word every tool uses and few define. Three rows and column names is a real limit you can check, not a reassurance.
When this is the wrong tool
Browser-native is a trade, not a free win:
- Very large files are bounded by your own RAM. A few hundred megabytes is comfortable on a normal laptop; tens of gigabytes is not, and a warehouse is the right answer there.
- Scheduled pipelines need a server. Nothing runs here while the tab is closed.
- Team-wide shared state is a database's job. Work is per-browser unless you sign in and save a dashboard deliberately.
What you can do with it
Load CSV, TSV or JSON; profile types and relationships automatically; clean and transform with live previews; query with SQL or plain English; build dashboards and charts; run correlation, regression and forecasting; and turn the result into a presentation. The analysis layer is complete — it is the storage layer that is different.
Try it on data that is not yours
The demo loads a sample sales dataset so you can see the whole path — raw rows to a written-up dashboard — before deciding whether to point it at anything sensitive.
Try it on sample data Open Malytics
No account needed to try it. The sample loads in your browser and nothing is uploaded.