AI Project Chief of Staff
Turn live project records into a focused list of interventions
- Difficulty
- Moderate
- Time to result
- ~weeks to results
- Steps
- 6
- Confidence
- 97%
This framework configures an AI assistant as a project-specific chief of staff. The leader supplies a compact set of living documents covering sprints, blockers, KPIs, roadmaps, and delivery dates. Each review asks the assistant to surface unresolved blockers, explain KPI gaps, identify overdue initiatives, and detect repeated patterns across reporting periods. The output is not merely a summary: it pairs each issue with its evidence, owner-provided reason, mitigation status, and proposed next action. The leader then uses the ranked findings to decide where personal intervention can unlock the most momentum. Consolidating each information category into one running document reduces context maintenance and makes the routine practical enough to repeat each morning or week.
Origin
Extracted from Marketing Against the Grain, where the host demonstrates the AI chief-of-staff routine used to review strategic projects and plan leadership attention.
Core principles
- 01Ground every recommendation in current project evidence
- 02Resolve blockers before they stall strategic momentum
- 03Treat missed KPIs as problems requiring reasons and recovery plans
- 04Look for recurring patterns rather than isolated misses
- 05End every review with accountable next actions
How to run it
- 1
Define the assistant's mandate
Tell the AI to act as a clear, concise, evidence-grounded chief of staff that challenges assumptions and ends with next steps.
Pro tip Explicitly require citations to the supplied records so every finding can be checked.
Watch out A vague role prompt can produce generic management advice rather than project-specific analysis.
- 2
Build a compact project context
Maintain running documents for sprints, KPIs, blockers, roadmaps, and delivery dates, then load the latest versions into the project workspace.
Pro tip Use one living document per information category to reduce repeated uploads.
Watch out Outdated records will produce outdated priorities.
- 3
Surface blockers
Ask for every unresolved blocker, its workstream, current mitigation, and any missing resolution plan.
Pro tip Set a policy such as resolving every blocker within the same week.
Watch out Do not accept a listed mitigation as proof that the blocker has actually been resolved.
- 4
Diagnose KPI gaps
Request missed metrics, the stated reason for each miss, the recovery plan, and whether the miss is isolated or recurring.
Pro tip Distinguish an emerging pattern from a metric missed repeatedly over several months.
Watch out Metrics without owners or explanations provide little basis for intervention.
- 5
Check schedule risk
Identify deliverables tracking behind their promised dates and connect each delay to its dependencies or blockers.
Pro tip Review delayed initiatives alongside KPI gaps because they may share a root cause.
Watch out Treating each delay independently can hide a systemic constraint.
- 6
Choose the highest-leverage action
Rank the findings by strategic impact and select the blocker or metric gap where leadership involvement can create the most momentum.
Pro tip Use the ranked agenda to plan the day or week rather than reacting to whichever message arrives first.
Watch out Do not let the AI make the final prioritization without applying business judgment.
In the wild
A project leader asks the assistant to summarize all blockers from July, indicate whether each has a mitigation plan, and propose one where none exists. The assistant groups findings by workstream and produces a summary table. The leader chooses the hardest high-impact blocker and schedules the necessary team intervention.
→ Leadership time is directed toward the constraint most likely to unlock project momentum.
The assistant reviews monthly KPI records and separates a metric missed twice from another missed four months in a row. It includes the reason and mitigation for each workstream rather than reporting only the latest values.
→ The leader can distinguish an emerging issue from a systemic performance problem.
Common mistakes
Using fragmented source documents
Uploading numerous overlapping files makes context maintenance difficult and increases the chance that the assistant uses stale information.
Reviewing misses without mitigation
A blocker or KPI gap should include an owner, explanation, and recovery plan; otherwise the review remains descriptive.
Delegating the final judgment
The assistant should clarify the evidence and options, while the leader remains responsible for deciding where to intervene.
Is it for you?
Best for
Leaders and project owners coordinating several workstreams, metrics, deadlines, and dependencies.
Not ideal for
Projects without reliable records or people willing to keep the underlying documents current.
From the transcript
“There's the blockers, there's the KPIs and then there's the KPI gaps and then there's any initiatives that are behind schedule.”
“I want to try to solve the hardest problems that can actually drive the most momentum”
“I set it up I give it all of the project details the project data the metrics the deliverables the road maps and I use…”
From the episode
How the Top 1% Use AI to 10x Their Productivity (Before Lunch)