The Missing UI for AI Data

Data quality is the most important factor in machine learning success. Hyperparam brings exploration and analysis of massive text datasets to the browser. Look At Your Data. 👀

Drop in a dataset (parquet, etc) to get started.

dataset.parquet

See what you're training on.
Drop in a dataset (parquet, etc) to get started.

What Makes Hyperparam Different?

Hyperparam is a new type of data app that can run entirely in the browser. This focus on frontend-first development (contrary to the python-centric data tools of the past) allows Hyperparam to be blazing fast and interactive.

Hyperparam is built from the ground up to handle modern data formats like parquet, and to scale to billions of rows. It can run entirely in the browser, or connect to a backend for larger datasets.

Time-to-first-data (lower is better)

155 ms

593 ms

862 ms

3466 ms

Hyparquet

Parquet WASM

Parquet JS

DuckDB WASM

Source: Hyparquet: The Quest for Instant Data

Model-Assisted Dataset Transformation

Welcome to the era of model-assisted data exploration and curation.

Hyperparam integrates large language models to help you understand and transform your data. Use natural language prompts to generate new columns, clean existing data, or summarize information. This makes it easy to prepare your datasets for machine learning tasks.

Live Demo: All of Wikipedia in the Hyperparam Viewer

This is a demo of what you can do with parquet files being read directly in the browser. A parquet file with all of the English Wikipedia articles is stored in S3 and rows are retrieved on demand using hyparquet.