Knowledge base
The internal document library — and the wider ground truth Olga reasons from.
Phase 1 + roadmap
"Knowledge base" means two related things in RunOlga. First, narrowly: the Knowledge tool — the internal library where organizational documents, procedures, and references live. Second, broadly: everything Olga grounds her answers in — the sum of the organization's records. Phase 1 ships both.
The Knowledge tool
Knowledge is the internal library for documents that don't belong to any single record — reference material, guides, procedures, anything the organization wants written down and findable.
What it holds:
- Documents organized by category
- For each document: title, category, content, owner, last-updated date
- Simple text search across all documents
How you use it:
- Browse by category
- View any document
- Create or edit documents — based on permissions
- Search across all documents
What Olga does here:
- Retrieves knowledge base content when relevant
- Refers to documents in conversations — "that's covered in the onboarding guide; here's the relevant part"
Documented Processes are close cousins: a Process is operational and owned, with a review cadence; a Knowledge document is reference material. When a new process needs documenting, it can start as a conversation with Olga and land in the Processes library.
The knowledge Olga grounds in
Beyond the Knowledge tool, every record the organization captures becomes part of what Olga can reason from: the Vision, the seats, the Priorities and Tasks, the open Issues, the documented Processes, and the Metrics. The more the organization writes down, the more grounded and useful Olga's answers become.
Two stores hold that knowledge:
| Store | What it holds |
|---|---|
| Instance Store (Layer 2) | The organization's content: its Vision, seats, Priorities, Issues, Processes, Metrics, Tasks, and Knowledge documents. Different for every organization. |
| Live Context Store | Current operational state, updated in near-real-time as people work in RunOlga. |
When you ask Olga a question, she retrieves the relevant records from these stores and answers from what she finds. She does not speculate and she does not fabricate facts about the organization. When she does not have the answer in her knowledge, she says so plainly. For how this composition works across all three layers — including her methodology layer — see how Olga works.
Every answer carries provenance: a record of which store it came from and which records were accessed. That makes Olga's knowledge traceable, not a black box. See grounding and confirmation.
You: What are our Priorities this quarter?
Olga: Here are the three Priorities on record for this quarter, each with its owner and target date.
This is retrieval, scoped to what you have permission to see. Olga reflects the knowledge back; she does not interpret it, rank it, or tell you what to do about it.
People knowledge
Roadmap — Phase 3
On the roadmap, the knowledge base extends beyond the operating records to what the organization knows about its people — a managed body of knowledge that supports people reviews and 1-on-1s and informs how leaders work with their seats.
With that knowledge in place, the intent is for Olga to help leaders prepare for decisions, issues, and reviews — to help the user decide based on the guidance the organization has captured. That is roadmap work, not Phase 1.
In Phase 1, Olga does not advise, recommend, or decide. She prepares, proposes, surfaces, and captures — and stops there. For exactly where the line sits, see Scope and roadmap.