Aeion Workspace as an Internal Company Wiki
Every internal wiki starts tidy and quietly rots into orphan pages, broken links, and answers nobody can find. Aeion Workspace fights entropy on purpose: `[[wiki links]]` with automatic backlinks, a knowledge graph that flags orphaned and broken pages, Tana-style supertags that auto-inherit fields, semantic search across five embedding providers, synced blocks for single-source content, and A–F health reports that tell you exactly what to fix first.
Why Internal Wikis Rot — and What Aeion Does About It
A flat folder tree hides its own decay. Pages go stale, links break, and knowledge quietly becomes an island nobody links to — until someone goes looking and can't find it. Aeion Workspace makes that structure visible and maintainable.
The Knowledge Graph — See the Shape of What You Know
Pages and records form a backlink/node graph you can visualize and navigate, rather than a flat list of folders.
Semantic Search — Find the Answer, Not Just the Keyword
People rarely remember the exact words a doc was written in. Semantic search across five embedding providers matches meaning, so the runbook surfaces even when the searcher's phrasing differs from the author's.
Graded Health Reports — A Checkup for Your Wiki
Instead of discovering stale pages and broken links by accident, get a report that grades the whole wiki and hands you a ranked to-do list.
Single Source of Truth — Synced Blocks and AI Q&A
Boilerplate that lives in ten pages goes stale in nine of them. Synced blocks fix that: a block of content is mirrored across many pages, and editing it once updates every instance — one source of truth for definitions, policies, or status snippets.
What This Wiki Is — And Isn't
Set expectations before you roll it out company-wide:
FAQ
Type `[[Page Name]]` and Aeion creates the link and automatically adds a backlink on the target page — no manual cross-referencing. Supertags go further: tag a page as, say, "Runbook" and it inherits that schema's fields instantly.
It's a force-directed graph over your pages' link structure that surfaces clusters, orphan pages (no incoming links), isolated pages, and broken links. It's deterministic graph analysis — it reads the links, not an LLM interpretation of the text.
Yes. Semantic search uses pgvector cosine similarity across five embedding providers, so pages match on meaning even without the exact keywords, typically in under 500 ms.
Yes — choose the local Ollama provider (all-MiniLM-L6-v2) and embeddings never leave your network. It's free and suited to HIPAA / GDPR-sensitive wikis. Provider is switchable per workspace.
Run the workspace health report. It grades content, tasks, collaboration, knowledge, and structure on an A–F scale, ranks the top issues (orphan pages, broken links, stale content), can run on a schedule, tracks trends week-over-week, and alerts the owner on a >10-point drop.
Synced blocks — a block mirrored across many pages that updates everywhere when edited once. Ideal for definitions, policies, and status snippets that appear in multiple docs.
Yes — the AI chat assistant does document Q&A against your own content, and per-column AI field agents can summarize or classify records automatically as they're saved.
Live over WebSocket for presence, cursors, and edits, but the current model is last-write-wins at the snapshot level — a real-time relay, not a CRDT engine, with offline-merge on the roadmap.
A wiki that stays findable as you grow.
Wiki links with auto-backlinks, a knowledge graph that flags orphans and broken links, semantic search across five embedding providers, synced blocks, and A–F health reports. $49/month flat, unlimited users.