Ingestion Upvoting

Ingestion Upvoting is the structured demand signal system that allows CueCrux users to influence which sources are prioritised for ingestion into the knowledge corpus. It creates a democratic mechanism for corpus development, ensuring that the platform's content grows in directions that serve the actual needs of its user community.

The upvoting interface appears in several contexts across the platform. When a user receives an answer with a thin coverage label, they are offered the option to suggest sources they believe would improve coverage. When browsing the Domain Directory, users can see pending sourcing requests and vote on the ones most relevant to their work. When submitting source suggestions through Community Threads, the suggestion automatically enters the upvoting queue.

Vote aggregation across tenants is where the system gains its power. A single user's vote on a source suggestion carries limited weight, but when multiple users across different organisations independently request the same content, the aggregated demand signal becomes a strong indicator that the content would be valuable to the broader community. This cross-tenant aggregation ensures that sourcing priorities reflect genuine widespread need rather than individual preferences.

Vote weighting adds nuance to the aggregation process. User trust scores, accumulated through accurate corrections and valuable community contributions, increase the weight of a user's votes. Topic relevance ensures that votes from users with demonstrated expertise in a domain carry more weight for sourcing requests in that domain. Subscription depth, reflecting how actively a user engages with a particular domain through the Watch Surface, adds additional context about the voter's commitment to and knowledge of the topic.

Priority scores computed from weighted votes feed directly into FactoryCrux's ingestion queue. High-priority source suggestions are processed ahead of lower-priority ones, with the ingestion pipeline allocating resources according to the community-determined priorities. The results of ingestion are visible in the Signals Feed, and users who voted for a source receive notifications when it is successfully ingested.

Transparency is maintained throughout the process. Users can see the current vote counts and priority rankings for all pending sourcing requests, understand how the prioritisation algorithm weights different factors, and track the status of sources they voted for from suggestion through ingestion to availability in the corpus.

For the CueCrux platform, Ingestion Upvoting transforms corpus development from a top-down editorial process into a community-driven effort where the users who depend on the platform's answers have a direct voice in determining what knowledge the platform acquires.