AutoML/In-browser/Zero upload

Train, tune, and read
your data—without it
ever leaving the browser.

Load a CSV, JSON, or SQLite file and DataWiz Pro profiles it, cleans it, selects features, and runs a Bayesian search across a dozen models for regression, classification, and clustering. Every byte of computation happens on your machine.

Open the workspace →See how it works
12
models tuned
TPE
Bayesian search
0 B
data uploaded
tuning · run_01 · regression
trial 00/24_
best R² · so far
0.61→ 0.947
leaderboard
  • ▸Random Forest0.947
  • 02Decision Tree0.911
  • 03k-Nearest Neighbors0.864
  • 04Ridge Regression0.812
✓ converged · 0 B uploaded

↑ a live tuning run, rendered from the same engine that powers the workspace

The flow

Six steps from raw file to a model you can trust.

Each stage hands a clean, inspectable result to the next. You can stop at any point and look at exactly what happened.

  1. 01

    Ingest

    Drop a CSV, JSON, or SQLite file. It parses in the tab — nothing is sent anywhere.

  2. 02

    Profile

    Types, distributions, missing values, and correlations are inferred automatically.

  3. 03

    Preprocess

    Impute, encode, and scale with a recommended pipeline you can override per column.

  4. 04

    Select

    Variance, correlation, and model-based scoring trim features down to the ones that matter.

  5. 05

    Tune

    A Tree-structured Parzen Estimator searches each model's hyperparameters under cross-validation.

  6. 06

    Read

    A ranked leaderboard and interactive charts explain what the best model learned.

What it does

An analytics team's toolkit, in a single tab.

no plugins · no setup

Ingest

Three ways in

  • ·CSV & TSV with delimiter detection
  • ·Nested JSON, flattened to columns
  • ·SQLite files queried with SQL
Model

A full roster

  • ·Ridge, k-NN, trees & forests
  • ·Logistic regression & naive Bayes
  • ·K-Means, DBSCAN & hierarchical
Tune

Bayesian search

  • ·Tree-structured Parzen Estimator
  • ·Cross-validated every trial
  • ·Live convergence as it runs
Clean

Honest preprocessing

  • ·Median / mode imputation
  • ·One-hot & ordinal encoding
  • ·IQR, z-score & Isolation Forest outliers
Read

Dashboards that explain

  • ·Leaderboard with full metrics
  • ·Feature importance & residuals
  • ·ROC, confusion & cluster maps
Keep

Your work, saved locally

  • ·Projects stored in the browser
  • ·Export a portable .dwproj file
  • ·Reopen exactly where you left off

Why in-browser matters

Your data never leaves your machine.

There is no upload step and no server to trust, because there is no server. Parsing, profiling, training, and tuning all run inside the browser tab. That makes DataWiz Pro a fit for regulated data, customer records, and anything you simply would not paste into a cloud tool.

  • ✓No accounts, no API keys, no telemetry
  • ✓Works offline once the page has loaded
  • ✓Projects persist locally and export as a single file
data path
▢ your browser
file→
profile→
clean→
select→
tune→
report
cloud servernever contacted

Pricing

The work is local. The plan scales with your team.

Every feature runs on every plan, because the engine lives in your browser. You pay for collaboration, support, and how you deploy it.

Individual

Free

The complete toolkit for one analyst.

  • ✓Every model and tuning method
  • ✓Unlimited local projects
  • ✓CSV, JSON & SQLite ingestion
  • ✓Export and import .dwproj files
Open the workspace

Team (CURRENTLY UNDER CONSTRUCTION !!)

Popular
$24/ user · month

Shared conventions for a working group.

  • ✓Everything in Individual
  • ✓Shared pipeline & report templates
  • ✓Branded report exports
  • ✓Priority support
Start a team trial

Enterprise (WOKRING ON THIS CURRENTLY !!)

Custom

Run it inside your own perimeter.

  • ✓Self-hosted internal distribution
  • ✓Security review & questionnaire support
  • ✓Volume licensing & SSO-gated hosting
  • ✓Onboarding and an SLA
Request a deployment guide