Serverless · In-browser · AI-assisted

Turn raw CSVs into
notebooks, tutorials & dashboards.

A complete scientific CSV intelligence platform. Automate cleaning, generate guided Python notebooks, and render interactive Chart.js dashboards — no server, no install, no code required.

Chart Types
0
Cleaning Steps
0
Notebook Cells
0
Models Ready
0runtimes
Scientific basis

Built on deterministic, reproducible primitives

Every cleaning pass, chart, and insight is derived from transparent rules so results are audit-ready for academic and enterprise workflows.

Deterministic cleaning

Header normalization, whitespace trim, duplicate removal, and null handling applied in a fixed, reviewable order.

Domain-aware analysis

A knowledge base engine turns known column names (medical, financial, jobs) into data-driven, contextual insights.

Reproducible notebooks

Every run produces the same .ipynb for the same inputs — version control, peer review, and replication ready.

Development explorer

Ship the app in 8 interactive steps

Walk through the build pipeline. Progress tracks as you click through each step.

Architecture explorer

Eight modules. One orchestrator.

Click any module to see the files it owns and the boundaries it respects.

Live studio

Three builders. Zero setup.

Everything runs in your browser — upload a CSV and generate real artifacts right now.

Evaluation

Benchmarked across three user profiles

Switch profiles to see how Notebook Studio performs in academic, SME, and enterprise contexts.

Walkthrough

Video overview

A 90-second tour of the entire cleaning → notebook → chart pipeline.

Claude Skill

Rebuild the whole app via Claude.

Download the portable datamentor-skill.md blueprint. Drop it into Claude Code or any Claude-powered tool and it recreates Notebook Studio end-to-end — cleaning pipeline, 10 charts, Pyodide notebook, Firebase auth, and this portal.

Source on GitHub