You upload a file. Everything after that is automatic.
You upload a file
Your file is read, not stored.
A CSV or Excel file, or a direct connection to a MySQL database. Additional database types are on the roadmap.
Free for files up to 500 rows and 20 columns — no account, no credit card. Larger files and database connections use a low-cost credit system.
It is read in memory to extract column statistics, never written to disk, and not retained after the assessment completes.
We take a sample of your file and mask any PII
AI never sees your raw data.
Rather than scanning the whole dataset, we draw rows proportionally across data categories, so no segment is over- or under-represented.
Sample size is calculated at 95% confidence and ±5% margin of error, scaled to your dataset. Minimum and maximum values from numeric columns are always included, so edge cases are never missed. Methodology developed in consultation with Bruce Ratner, PhD, Predictive Analytics Consultant.
A regex layer catches structural patterns — email, SSN, credit card, phone. Google Cloud Sensitive Data Protection catches semantic PII regex alone misses: person names, street addresses, dates of birth. Detected values are replaced with placeholder tokens such as [PII:EMAIL] and [DLP:PERSON_NAME].
Column names, data types, null rates and sample values go to OpenAI — never your raw data. OpenAI never sees your original PII values. Masking runs on our servers before AI analysis begins.
We analyze and score your data
Every column read — not just its name.
For every column we examine the actual data, not the column name alone, to determine what it really contains and resolve ambiguity where possible.
Four dimensions, scored per column.
Columns are analyzed in parallel, so assessment time barely moves with dataset size. A measured run on an 18-column file takes about 15 seconds from selecting the file to the score on screen.
We deliver the results
Four things you can act on.
A score from 0–100 placing your dataset from Not AI Ready to AI Ready, with specific findings for every column.
Concrete, ordered actions built from your data — which columns to rename, where nulls will distort an answer, what to mask — before AI sees it.
Column names, detected types, quality scores, null rates, ambiguity flags, PII detection and semantic descriptions. Exportable to Excel for team sharing, onboarding and data governance documentation.
| Column | Type | Quality | Null rate |
|---|---|---|---|
| src_cd | text | 10 | 9.7% |
| sess_dur | numeric | 10 | 11.8% |
| emailPII | text | 10 | 63.9% |
Example output — Web Session Data sample dataset.
Five ready-to-paste prompts with the dataset's context already inside them, so you don't have to explain your data before you can ask it a question.
The semantic layer your AI tool is missing — a draft definition of what each column contains and how AI should interpret it, including what null means in a column with a high null rate, which is where AI tools that write their own queries silently return incomplete answers. Review and edit before use.
Results are directionally accurate for identifying data quality and readiness issues. The AI Data Maturity Score is a practical assessment guide — not a formal statistical audit. For datasets where precision matters, we recommend reviewing flagged columns manually.
Free to start — no account, no credit card, for files up to 500 rows / 20 columns.
Assess My Data →