kasipa
Back to blog

Kasipa: AI-powered data analysis for everyone

· Dylan

Data analysis shouldn't require a computer science degree. Yet for most people, the gap between having data and actually understanding it remains painfully wide. Spreadsheets are powerful but intimidating. Python notebooks are flexible but require coding skills. BI tools are polished but expensive and complex.

That's why we built Kasipa.

What is Kasipa?

Kasipa is an AI-powered data analysis platform. You upload a CSV, Excel file, or connect a Google Sheet, then ask questions in plain English. The AI agent writes and executes Python code on your behalf — right in the browser — and returns answers, tables, and interactive charts.

No installs. No API keys to manage. No code to write.

How it works

The workflow is simple:

  1. Upload your data — drag and drop a CSV or Excel file, or paste a Google Sheets link
  2. Ask a question — "What are the top 10 products by revenue?" or "Show me the trend of monthly signups"
  3. Get results — the AI writes the required code, executes it in your browser, and displays the output

Everything runs client-side using Pyodide, so your data never leaves your browser. You can even use your own model keys for more control.

What you can do

Here are a few examples of things you can ask:

  • "Summarize this dataset" — get row counts, column types, and basic statistics
  • "Show me a bar chart of sales by region" — interactive Plotly visualizations
  • "Filter rows where status is active and revenue > 1000" — transform your data
  • "What's the correlation between price and rating?" — statistical analysis
  • "Clean up the date column and remove duplicates" — data wrangling

The AI understands context, so you can ask follow-up questions and build on previous results.

What's next

This is just the beginning. We're working on:

  • More data connectors — databases, APIs, and more
  • Collaboration features — share analyses with your team
  • Scheduled reports — automate recurring analyses

We'd love your feedback. Try it out and let us know what you think at [email protected].