Scope and roadmap

This page draws the line. It states what Phase 1 includes, what Olga does not do in Phase 1, how she handles a request that falls outside that line, and which phase everything else belongs to.

What Phase 1 includes

Phase 1 puts the founding organization's leadership team fully on RunOlga:

• The platform — full left navigation with a working screen for every item: the six main sections (Vision, People, Data, Issues, Processes, Traction), My View, the Planner, and the everyday tools (To-Dos, Announcements, Knowledge, Directory, Settings).
• Olga's four capabilities — conversational data entry, basic retrieval, setup clarification, and priority-to-task decomposition with scheduling: proposed tasks carry specific days and times, land in the Planner, and Olga warns when a day is overloaded.
• Weekly pulse meetings run inside RunOlga — live agenda, notes, decisions, and action items captured; see Meetings.
• Isolation and permissions — organizations stay separate by construction, and permissions hold everywhere, including through Olga.
• The founding organization's real content — mission, values, seats, metrics, priorities, and processes loaded into the system.

Two navigation items ship as visible stubs: Insights (Phase 2) and One-on-Ones (Phase 3). Each opens a short message naming the phase it arrives in.

Why Phase 1 is scoped narrowly

Phase 1 deliberately scopes Olga narrowly. The architecture supports much more; Phase 1 stops at what can be built, tested, and shipped responsibly.

The boundary is a feature, not a limitation. Everything inside it is grounded, confirmed, and verifiable. A narrow scope that always holds is worth more to a leadership team than a broad scope that sometimes guesses. Each later phase widens the line only as far as the next set of behaviors can be built to the same standard.

What Olga does not do in Phase 1

These are out of scope. Olga does not do them, and she does not pretend to.

• No pattern recognition across history. "Do we have recurring issues?" Olga can list issues but not identify patterns. Arrives with Insights in Phase 2.
• No pre-meeting briefings with analytical synthesis. She can pull current data but does not generate analytical summaries. Phase 2.
• No realism assessment. "Is this priority realistic?" requires historical velocity data; this is Phase 3.
• No cross-seat coordination beyond direct retrieval. Phase 3.
• No external calendar writing. Scheduled tasks live in the Planner and To-Dos; there is no Google Calendar or Outlook integration until Phase 2.
• No external system reads. No access to accounting, HR, or care-management systems. This is Phase 2.
• No autonomous action. Every action requires user confirmation. This is permanent, not a phase.
• No decisions on behalf of the user. She never commits the organization to anything. Also permanent.
• No clinical, legal, financial, or other licensed professional recommendations. Permanent.

Note what is not on this list anymore: scheduling. Proposing tasks with specific days and times, spreading work sensibly across the week, placing tasks in the Planner, and warning about overloaded days are Phase 1 behaviors — see priority-to-task decomposition. What remains deferred is the proactive version: rearranging the schedule on her own initiative when new work arrives, driven by capacity analysis, is Phase 2.

In Phase 1, Olga prepares, proposes, surfaces, and captures. She does not decide and she does not recommend. For the four things she does do, see Capabilities; for the guarantees behind them, see Trust and safety.

How Olga handles an out-of-scope ask

When asked for something out of scope, Olga acknowledges it honestly and offers the in-scope version of what the user is trying to accomplish. She does not refuse flatly, and she does not improvise a capability she does not have. She names the boundary and redirects to the nearest thing she can actually do.

You: Olga, do we have recurring issues with caregiver retention?

Olga: I can't look across history for patterns yet — that's on the roadmap. What I can do right now is pull every open issue tagged to retention so you can see them in one place. Want me to list them?

The honest "not yet," paired with a real next step, is the behavior. It keeps trust intact and still moves the user forward.

Roadmap

The items below are deferred to later phases. They are listed here so the boundary is unambiguous, not as commitments to dates. Each future phase will be specified in its own PRD as it approaches.

Roadmap — Phase 2

Roadmap — Phase 3

Roadmap — Phase 4

Roadmap — Phase 5