For AI labs

Real work data from the companies other programs skip.

SellMyData lists de-identified business data (messages, code, documents, records and tickets) from small and mid-size companies. Every dataset is approved by the company that owns it.

What’s on the exchange.

01

Messages

Threads where teams coordinate work and make decisions.

Slack · Teams · Gmail

02

Code

Repositories with the commits, reviews and issues behind them.

GitHub · GitLab · Jira

03

Documents

Contracts, SOPs, reports and the written record of how a company operates.

Drive · Notion · SharePoint

04

Records

Pipelines, invoices, ledgers and orders from the systems that run the business.

Salesforce · HubSpot · QuickBooks

05

Tasks

Tickets captured end to end: the brief, the back-and-forth, the result.

Zendesk · ServiceTitan · Asana

Request it. Preview it.
License it.

  1. STEP 01

    Post a request

    Tell us the data type, industry and years of history you need. Open requests appear on the demand board.

  2. STEP 02

    Preview redacted samples

    See a de-identified sample of each matching dataset before you buy.

  3. STEP 03

    Agree the license

    The company that owns the data approves the buyer and the terms before anything is licensed.

  4. STEP 04

    Receive the dataset

    You get the de-identified dataset, never the company’s raw data.

The long tail of real work.

Most data programs only work with large companies. The record of how work gets done at a 12-person firm (dispatch threads, month-end close, support queues) rarely reaches a training set.

SellMyData is built for those companies: no minimum headcount and no sales call to list. Scope, format and delivery are agreed for each dataset, so tell us what your model needs and we’ll tell you what’s available.