Send one complete dataset brief and give reviewers enough information to understand the product before opening the sample. A clear submission can move into directory review without a long email chain.

We review useful business, research, publishing, and AI datasets in formats such as JSON, CSV, JSONL, and Parquet. The creator must have the right to share or license the records.

Prepare the Dataset Review Package

Name the buyer and job. Explain who needs the data and what it helps them decide or build.
Document the records. Include the source, collection method, coverage, update date, format, and field definitions.
Show a safe sample. Attach sample records with no private personal, health, payment, or account information.
State the deal. Include the license, allowed uses, known limits, delivery method, price model, and official source URL.

Send This Structured Email Brief

Send the brief to submit@aitoaidatasets.com with the subject Dataset Directory Submission. Use these headings in the message:

Dataset name: the public name buyers will see
Publisher: the person or company responsible for it
Buyer and use: who needs it and what job it supports
Source and rights: where the records came from and why you may share them
Coverage: dates, locations, subjects, and important gaps
Format and sample: file type, schema, and sample link or attachment
Quality and limits: missing values, duplicate handling, validation, and known weak spots
License and access: permitted uses, price model, delivery, updates, and official URL

A listing may identify an answer as unknown when the rest of the submission is useful. Missing information will not be turned into a claim.

What We Do Not Accept

Data the sender does not have the right to share
Private personal information or protected health information
Samples that expose payment, login, account, or security details
Files with no source, purpose, or usable field documentation

Make the Offer Easier to Review

Use the Sell-Ready Dataset Worksheet before sending the brief. If you are evaluating somebody else's collection, use the AI Dataset Buyer Scorecard instead.