The Insights Graph visualizes your knowledge as an interactive network. Every document, note, or conversation you ingest gets broken into concepts (entities) and their connections (relationships), then rendered as a force-directed graph.

Node size = importance (PageRank). Bigger nodes are more connected concepts.

Node color = community. Nodes clustered together share related topics (Louvain detection).

Edges = relationships between concepts. Thicker lines mean stronger connections.

Dashed lines = cross-community bridges. These connect different topic clusters.

Hybrid — Combines vector similarity + graph traversal. Best quality.

Vector — Pure semantic similarity search. Fast.

Graph — Follows relationships between entities. Best for exploring connections.

Use the source pills at the top to filter by data source. Connect additional sources (X, Reddit, GitHub, etc.) in Settings to see cross-platform insights.

Click any node to see its details and connections.

Use + Ingest to add documents, notes, or articles.

Use Insights to see AI-detected knowledge gaps between communities.