Second Data
Data types

Real work, recorded over years.

The most valuable data shows how a business actually runs: the questions customers ask, the way teams solve problems, the code and the reviews behind it. Here is what we look for.

Personal details removed before training

Personal and confidential details are removed before any AI team trains on the data: names, email addresses, phone numbers, account numbers, client names, and passwords or keys. Every license requires this, and forbids any attempt to re-identify people.

Team communications

High demand

Workspace exports that show how people work together and make decisions.

  • Slack, Microsoft Teams, and Discord workspaces
  • Internal email archives
  • Meeting notes and recorded standups

Customer support

High demand

Full conversation histories between customers and your team, including how each issue was resolved.

  • Zendesk, Intercom, Freshdesk, and Help Scout tickets
  • Live chat and email support logs
  • Help center articles and internal macros

Code and engineering

High demand

Private codebases with their history, plus the discussion around them.

  • Git repositories with full commit history
  • Pull request reviews and comments
  • Jira, Linear, and GitHub issues
  • Incident reports and postmortems

Documents and wikis

The written knowledge of the company.

  • Notion, Confluence, and Google Drive
  • Process documents, playbooks, and training material
  • Proposals, RFP responses, and reports

Sales records

How your team found, pitched, and closed customers.

  • CRM activity and notes (Salesforce, HubSpot)
  • Call recordings and transcripts
  • Outbound sequences and replies

Operations and finance

Structured records from day-to-day operations.

  • Invoices, receivables, and payment histories
  • Work orders, field reports, and inspection records
  • Logistics, inventory, and scheduling data

Specialist expertise

Work product from skilled professionals in a specific field. These archives are rare and often command the highest prices.

  • Engineering, legal, finance, and accounting work product
  • Scientific and lab records
  • Technical documentation and specifications

What makes data worth more

  • Volume. More records, more users, more activity.
  • Time span. Several years of history beats several months.
  • Completeness. Whole threads and full histories, not fragments.
  • Specialization. Expert work in a specific field.
  • Clean rights. Clear ownership and no contract limits on licensing.

What we do not take on

  • Data you do not have the right to license
  • Payment card numbers, bank credentials, or government ID numbers
  • Protected health information
  • Data collected from children
  • Data obtained by scraping or without permission

Have something that is not on this list?

Tell us about it. Buyer demand changes fast, and unusual archives are often the most valuable.

Get a free valuation