The Living Execution Plan
Each Execution Scope has one Living Execution Plan — the shared, continuously-updated view of what the scope is actually doing, who owns what, and what matters right now.
What it is
In Parallel maintains one Living Execution Plan per Execution Scope. It exists to answer the questions that teams constantly ask — and often can't answer quickly:
What are we actually doing right now?
What matters most?
Who owns what?
What's at risk?
What changed recently, and why?
Unlike a static document or a project board, the plan is designed to stay truthful as work evolves. It updates from meetings, connected systems, and your review — not from someone manually rewriting it.
What's in the plan
The plan organizes execution reality in a hierarchy: Goals → Priorities → Activities, with obstacles and owners attached at each level. Goals anchor direction. Priorities rank what matters most. Activities are the concrete work being executed.
Two principles keep the plan legible at scale:
The 30-second rule — the plan's current state should be readable in under 30 seconds. If it can't be, the scope is carrying too much detail or is too broad.
Top 3 priorities — the plan surfaces the three highest-priority items prominently. This forces real ranking rather than everything becoming equally urgent.
Section | What it shows |
Goals and commitments | The outcomes the scope is accountable for — anchors the plan so priorities don't drift into activity without direction |
Priorities | What matters most right now, ranked — forces tradeoffs and prevents everything from becoming urgent |
Tasks and milestones | Concrete work and checkpoints at a high-signal level — delivery detail stays in Jira/Asana/etc. |
Risks and dependencies | Blockers, constraints, cross-team dependencies, and confidence issues |
Ownership | Who is accountable for each area — explicit, never implied |
Recent changes | What shifted in the plan and when — so change is visible and reviewable rather than silently overwritten |
Observation widgets on the dashboard have an arrow (→) in the widget header that links directly to the relevant Findings list for that observation type.
How it stays up to date
Three mechanisms keep the plan current:
Meeting Intelligence Pipeline — the Transcriber captures signals from each connected meeting; observations become structured proposals reviewed and confirmed after each session
Connected tools — your calendar and connected communication tools feed signals into the plan
Drift Detection — the AI compares stated priorities against observed activity and surfaces divergence before it compounds into a larger problem
Meetings produce signals. When the Transcriber joins a connected recurring meeting, it captures decisions, actions, risks, and ownership shifts — and links them to the scope as execution signals, not meeting notes.
Reports turn signals into structured outputs. After each meeting, a post-meeting report is published. You review it, confirm what matters, and the plan updates. Tasks are assigned and stakeholders can be notified.
Human review keeps it trustworthy. Nothing material changes silently — every plan update goes through a review and confirmation step. This is what makes the plan durable: it's a traceable evolution, not whatever someone last edited.
Weekly review habit
A strong weekly review takes minutes and prevents drift.
Review what changed. Scan what moved in priority, what changed in risk, what shifted in ownership, and what new actions appeared since the last session.
Confirm priorities still match reality. Ask: are we working on the right things? Is the ranking honest? Has anything become urgent that isn't reflected yet?
Scan risks and dependencies. Identify anything that needs a decision, cross-team coordination, or escalation.
Close loops. Confirm that actions have owners, blocked items are explicitly marked (not silently stalled), and completed items are closed.
Tip: If you review the plan weekly and confirm after key meetings, clarity compounds quickly — the plan becomes the default answer to "what's happening?" rather than something you reconstruct each time.
Related