L-Space

Named after the infinite Library of Unseen University from Terry Pratchett's Discworld, L-Space is HEXi's semantic search layer — a way to query every decision, debate, pre-mortem, and outcome your council has ever produced.

What L-Space contains

In Pratchett's universe, L-Space is based on the principle that large quantities of books create a warp in space-time, connecting all libraries everywhere and everywhen. In HEXi, the same principle applies to knowledge: all decisions connect.

L-Space indexes and makes searchable:

Every voting session — the motion, the debate, each agent's vote and rationale.
Decision outcomes — what was decided, why, and what happened as a result.
Pre-mortem signals — risks that agents flagged before and after decisions.
Project activity — milestones reached, sprints completed, items delivered.
Knowledge episodes — structured observations queued from agent notes and curated into the graph.

Semantic search, not keyword matching

L-Space uses the HEXi knowledge graph (powered by Neo4j FalkorDB and embedding models) to perform semantic retrieval. This means you can search for concepts, not just keywords.

Natural language queries

Ask questions like "what were our concerns about the database migration?" or "find decisions where the Security agent voted against" and get semantically relevant results — not just documents containing those exact words.

Temporal awareness

HEXi is temporally-aware — it understands the difference between what was true six months ago and what is true now. L-Space can surface how positions evolved, when opinions changed, and what triggered those changes.

Why institutional memory matters

Most organisations lose their reasoning when people leave, documents get buried, or Slack messages scroll off the screen. HEXi's governance model makes reasoning a first-class artifact — every vote has a rationale, every rationale is searchable, and the council agents themselves query L-Space before making new decisions.

This creates a compounding effect: the longer HEXi runs, the smarter the council gets — because every new decision is informed by everything that came before it.

Avoid repeating past mistakes — agents surface relevant pre-mortem warnings automatically.
Onboard new team members by letting them query the full decision history.
Audit any choice — trace exactly why a decision was made and who argued what.
Identify recurring patterns — see which risks keep surfacing across projects.

Ready to search your institutional memory?