For AI labs and data buyers

Real operating data, licensed and ready to train on.

We source proprietary data from operating companies across industries: support conversations, code, docs, CRM and operations records. De-identified, and licensed with the seller’s approval.

Request data

What’s available

How real work gets done, from request to result.

Support tickets

Real customer problems and how experienced teams resolved them.

Source code

Internal tools and services, with the commit history that explains them.

Docs & SOPs

Runbooks, onboarding guides and how-tos from working teams.

CRM & sales notes

Call notes, objections and deal reviews, with contacts removed.

Chat & email

How decisions get made inside companies, with identities stripped out.

Ops data & logs

Telemetry paired with the notes that explain what happened.

How it works

Three steps

  1. 01Tell us what you needDomain, data types, volume and timeline. We reply within one business day.
  2. 02Review redacted samplesSee representative, de-identified samples before you commit to anything.
  3. 03License with clear termsA signed license that names the use, the term and when raw data is deleted.

Every dataset comes with

Clean, cleared, documented

  • De-identifiedNames, contact details, account numbers and secrets in code are removed before any sample.
  • Seller approval on recordEvery license is approved by the company that owns the data, for your stated use.
  • Clear provenanceWhich kind of company, which systems and which years the data covers.
  • Delivery that fitsFiles, a bucket or an API, in the formats you already work with.

Request data

Tell us what you’re looking for.

We reply within one business day with what matches.

Timeline