Start with the information need
An AI-assisted research process may need current public page content, source links or a defined set of factual observations. Identifying that need comes before deciding how the information should be collected.
Make provenance part of the brief
Source URLs and collection times can matter as much as the extracted text. They help a receiving workflow distinguish an observation from an inference and understand when the underlying source may have changed.
Discuss the receiving workflow
Tell us how the information will be consumed. We can discuss the collection and preparation requirements without assuming a particular model, agent framework or output schema.
- 01Define the public sources
- 02Identify the required information
- 03Discuss freshness and source context
- 04Agree how the receiving system will use it
Respect the boundary between access and use
Public availability does not remove every restriction on reuse. Source terms, intellectual property and privacy need to be considered for the intended application. This website does not advertise model hosting or a ready-made AI application.