Contributing Knowledge

Contribute your work to a databases, enriching your organization's knowledge

This guide explains how to use Insights to contribute durable knowledge to your organization’s knowledge base. Insights capture verified takeaways from conversations with provenance, making them ideal for publishing, indexing, and reuse across teams and systems.

Why contribute with Insights

  • Trustworthy summaries: Curated, human-readable outputs with links back to sources and messages.

  • Reusable artifacts: Portable across tools and teams (e.g., CRM, wiki, reporting, RAG indices).

  • Governed access: Inherit conversation/bot permissions and sharing rules.

What you contribute

  • Titles and summaries that encapsulate the knowledge.

  • Provenance via message excerpts and document references.

  • Contextual metadata (owner, timestamps, model used, curation flags).

Contribution workflow

  1. Generate or curate insights in a conversation

    • Use your agent/bot to synthesize key findings; refine as needed.

    • Select which insights should be connected to the data room.

  2. Review selected insights

    • Validate title, summary, and provenance.

  3. Enrich and normalize

    • Ensure the summary is self-contained and actionable.

    • Map metadata (owners, tags, categories) to your knowledge base schema.

  4. Publish to knowledge destinations

    • Export to a wiki/portal (e.g., create pages with summary + references).

    • Index into search/RAG systems (store summary, embeddings, and doc refs).

  5. Maintain and update

    • Pin, edit, or re-generate insights as understanding evolves.

    • Replace expired document refs by re-fetching insight details.

Last updated

Was this helpful?