01
ACTIVE WORKSPACEHome

Purpose, privacy model, outputs, and the EDA workflow.

LOCAL-FIRST EXPERIENCE | CLOUDFLARE PAGES READYREVIEW AUDIO READY

Understand tabular data before building a model. schema inference|

A privacy-aware browser workbench for first-pass exploratory data analysis, schema inspection, missingness checks, summary statistics, and portable profiling reports.

Original template
user@metricforge: ~/vite-data-science-workbench


LocalCSV processingAutotype inferenceJSONprofile export0uploads










01 | CLEAR PRODUCT DEFINITION
Useful by design, honest about scope.
Profile tabular data locally so structure and quality issues are visible before feature engineering or model training begins.




Primary use
First-pass EDA, dataset quality checks, statistics teaching, and lightweight portfolio demonstrations.
A privacy-aware browser workbench for first-pass exploratory data analysis, schema inspection, missingness checks, summary statistics, and portable profiling reports.
GitHub repository ↗


Produces
Concrete, reusable outputs
Column types, missingness, unique counts, summary statistics, quality charts, and JSON profiles.
Boundary
No exaggerated capability claims
Browser memory and JavaScript numeric limits make it unsuitable for very large datasets or advanced modeling without dedicated compute.










02 | WORKFLOW
One focused path from input to output.
Every stage is separated so the interface stays readable, reviewable, and useful for AI and Data Science learning.
01
Load
Paste CSV text or choose a local CSV file.
02
Infer
Detect columns, basic data types, missing values, and unique counts.
03
Summarize
Calculate descriptive statistics and quality indicators.
04
Export
Download a structured JSON profile for documentation or follow-up work.