What business data can be monetized?

Software and its development context

A codebase can show implementation in a real domain. Tests, build instructions and architecture notes can make it easier to understand. Commit history, linked issues and reviews can add the context of how changes were made.

Do not assume all repository content belongs to the company. Dependencies, copied snippets, external contributions and contractor work need a rights inventory. Archived software may still have relevant structure, but discontinued does not mean free of obligations.

Workflows with recorded outcomes

Process records can be more interpretable when they link a request, a decision, an action and an outcome. Examples might include a documented maintenance process or a quality-control workflow; these are illustrative categories, not confirmed buyer demand.

Customer tickets, employee evaluations and transaction records can contain personal or confidential information. A business purpose for collecting them does not automatically authorise licensing for AI use. Consider a narrower, safer collection and a legal review before discussing samples.

Documentation and specialist datasets

Specifications, runbooks and technical manuals can explain systems and procedures. Industry datasets may include structured measurements or domain records with meaningful labels and coverage. Documentation of origin, format and limitations is important for interpretation.

Check whether contracts permit sharing and whether underlying material comes from suppliers, customers or public sources. Purchased or accessible data may be usable internally without being licensable onward. Avoid mixing sources when you cannot explain their permissions.

Screen candidates with a simple inventory

Create one row per collection rather than cataloguing every file. Record the responsible team, format, coverage period, approximate size, how it was collected and why it could be relevant to an outside task. Add a separate rights and sensitivity note.

Uncertainty is a reason to investigate, not a reason to invent clearance. AssetGate’s assessment flags uncertain rights and personal data for review rather than automatically rejecting the description.

• What is it, and which business process produced it?
• Can you connect the material to context or outcomes?
• Who owns it, and which third-party agreements apply?
• Does it include personal, customer-owned or regulated information?
• What could be excluded without losing the useful structure?

What not to send

Do not upload raw customer records, production databases, credentials or confidential source code to obtain an initial assessment. Descriptions are enough to start. Avoid including individual names or sensitive examples in free-text fields.

If later interest justifies a scoped review, agree permission, privacy safeguards and access terms first. Removing direct identifiers does not by itself prove anonymisation, and an NDA cannot replace the necessary rights.

Choose a low-risk first step

Prioritise a collection you can describe accurately, whose origin is documented and whose restrictions are understood. Keep preparation proportionate until there is a credible use case and genuine interest.

The free seller assessment provides qualitative guidance only. It does not establish a sale price or guarantee that an AI buyer will want the material. You remain in control of whether to continue.