Parquet AI

Upload your Parquet file and interact with it using natural language. Your data stays private and secure on your device.

PrivateFreeNo Signup

Quick Start

  1. 1
    Upload your Parquet file

    Drag and drop a Parquet file into the upload area. The file is processed entirely in your browser.

  2. 2
    Ask a question in natural language

    Type your question about the data. Llama-3-70b AI generates SQL based on your natural language query.

  3. 3
    View results instantly

    The SQL is executed in your browser directly against the Parquet file without data leaving your device.

Specs

AI Model
Llama-3-70b
Processing
WebAssembly (in-browser)
Query Language
SQL (auto-generated)

Data stays on your device

Your Parquet data never leaves your device. Only the column schema is sent to generate the SQL query.

Parquet AI leverages WebAssembly and Llama-3-70b AI to let you query your Parquet files using natural language. It detects column types and names, generates SQL from your questions, and executes it directly in your browser — no data ever leaves your device.

Llama-3-70b

AI Model

In-Browser SQL

Execution

100% Client-Side

Privacy

Natural Language Queries

Ask questions in plain English and get SQL generated automatically by Llama-3-70b AI.

In-Browser Execution

SQL is executed locally using WebAssembly — your Parquet data never leaves your device.

Smart Schema Detection

Automatically detects column types and names in your Parquet file to generate accurate queries.

How does Parquet AI generate queries?

Our tool detects column types and names in your Parquet file, then prompts Llama-3-70b AI to generate SQL based on your natural language query. The SQL is executed in your browser using WebAssembly.

Does my data leave my device?

Your Parquet data never leaves your device. All processing and SQL execution happens locally in your browser using WebAssembly. Only the column schema is sent to generate the SQL query.

What kinds of questions can I ask?

You can ask any question that can be answered with SQL — filtering, aggregation, sorting, grouping, and more. For example: 'What are the top 10 rows by revenue?' or 'How many unique categories are there?'