EazyDataFix v2.0.0 · Reusable preparation recipes

Messy files in.Decision-ready data out.

Define preparation once, review every proposed change, then replay the same rules on future CSV or Excel files for analysis, machine learning and Power BI.

Open Data Studio View launch pricing

Deterministic by default Complete change log AI does not control fixes

Analysis Ready workflow

Transparent preview · no black box

90 → 100

customer_upload.csv

3 rows shown

customer_id	name	salary	city
C-101	Anika Rao	₹72,000	HYD
C-102	Rahul	N/A	Hyderabad
C-102	Rahul	N/A	hyderabad

6 reviewable changes

Ready for analysis

2 rows shown

customer_id	name	salary	city
C-101	Anika Rao	72000	Hyderabad
C-102	Rahul	null	Hyderabad

v2.0.0

Latest stable release

3.10–3.13

Python verified

JSON · YAML

Versioned recipes

MIT

Open-source core

New in v2.0.0

Turn one careful cleanup into a reusable pipeline.

Save explicit column types, mappings and conversion policies as a versioned recipe. Preview the result, review selected stages and apply only the output you inspected.

JSON or YAML recipe replay

Preserve, reject or quarantine invalid values

Source and recipe integrity checks

Shared evidence across Analysis, ML and Power BI

Read the recipe guide View release notes

monthly_sales.json

Portable rules · deterministic replay

recipe = edf.PipelineRecipe(
    name="monthly_sales",
    columns={
        "customer_id": edf.ColumnRule(dtype="string"),
        "amount": edf.ColumnRule(
            dtype="number", currency_symbols=("₹",)
        ),
        "order_date": edf.ColumnRule(
            dtype="datetime", date_format="%d/%m/%Y"
        ),
    },
)

pipeline = edf.Pipeline(recipe)
plan = pipeline.preview("sales.csv")
result = pipeline.apply("sales.csv", plan)

One engine · three outcomes

Prepare data for the work you are actually doing.

Generic cleaning is not enough. EazyDataFix applies different readiness checks for analysis, supervised machine learning and Power BI models.

Analysis Ready

Stop cleaning before every analysis

Standardize fields, handle missing markers, remove configured duplicates and export a typed dataset with readiness evidence.

Clean column names
Safe type conversion
Quality score + report

ML Ready

Prepare features without silent leakage

Create train/test inputs with transformations fitted only on training data and a reusable preprocessing artifact.

Target-aware preparation
Leakage safeguards
Reusable artifact

Power BI Ready

Give Power BI cleaner model inputs

Validate field types, keys and relationships, generate a date table and export documented model-ready files.

Key validation
Relationship checks
CSV/Excel export pack

From upload to evidence

A workflow people can trust, not an “auto-fix” button.

EazyDataFix is designed around inspection. The customer sees the problem, the proposed action and the resulting dataset before exporting anything.

Try the workflow

Bring a real file

Start with the CSV or Excel file already slowing down your reporting or analysis.

See what is wrong

Review missing values, duplicate records, unsafe types, inconsistent labels and model-readiness issues.

Approve every change

Accept or reject proposed corrections. Important data is never silently rewritten.

Export with evidence

Download clean data together with scores, warnings, validations and a complete change log.

Trust is a product feature

Deterministic first. AI optional and accountable.

Cleaning decisions come from reproducible rules—not a model guessing what your data should mean. Optional AI can explain evidence after the deterministic workflow, but it does not secretly rewrite source values.

Preview before applying

Protect identifiers and leading zeroes

Document every transformation

Keep the open-source engine inspectable

Built on an open-source Python engine

Use the Studio or own the code path.

import eazydatafix as edf

recipe = edf.PipelineRecipe.load("monthly_sales.json")
pipeline = edf.Pipeline(recipe)
plan = pipeline.preview("sales.csv")
result = pipeline.apply("sales.csv", plan)

if result.is_ready:
    result.save("sales_output")

Developer documentation GitHub

Start with the file that is wasting your time today.

The public Studio preview runs locally in your browser and demonstrates the complete review workflow. Assisted team pilots add repeatable recipes and business-specific rules.

Try Data Studio Compare launch plans