Show notes
From Novice Prompts to Expert Systems: How to Build AI Workflows That Run Your Routine Work While You LeadIn Part 1, we explored the foundational mindset for AI executive productivity—the shift from 80% routine work to 80% creative work. Now, in Part 2, Barry O'Reilly reveals the specific AI executive workflow automation systems that make this transformation real.This isn't theory. These are the exact workflows, prompts, and systems that Barry and leading executives use daily to reclaim their time and amplify their leadership impact.What you'll learn: The weekly business review system that takes 90 seconds instead of 30 minutes. How to audit your work with AI's help. Building your personal prompt library. And why your unlearning rate must exceed your irrelevance rate.The AI Executive Workflow Automation Philosophy: Creative Work + Automated DisciplineBefore diving into specific systems, understand the core principle driving effective AI executive workflow automation:
"Every time you can automate routine but disciplined work, you're moving the needle toward having more capacity to do creative problem-solving work. That's where you get the power and real promise of what AI is—people doing the best work of their life."
Barry O'Reilly
The Work Category FrameworkAI executive workflow automation works by understanding two distinct categories of leadership work:Category 1: Creative Problem-Solving WorkStrategic planning and vision developmentComplex decision-making in ambiguous situationsCoaching team members through challengesInnovation and new product/service designBuilding relationships and influencing stakeholdersPattern recognition across diverse business situationsCategory 2: Routine Disciplined WorkWriting meeting follow-ups and summariesTracking action items and deadlinesSending reminder notificationsCompiling weekly/monthly reportsScheduling and calendar managementData entry and information organizationThe AI Executive Workflow Automation Insight: Humans should do Category 1. Machines should automate Category 2. The problem is most executives spend 80% of their time in Category 2.As Barry explains: "Machines essentially offer this opportunity to automate a lot of that disciplined, repeatable, routine work—like having an auto-scheduler that sends an email 5 days before a task is due. I don't want to think about it, I don't want to send it, but a machine is amazing at making sure it follows up and does that."The Self-Audit: Ask AI to Analyze Your Work EfficiencyThe first AI executive workflow automation you should implement is having AI audit where you're spending your time. This creates objective data about your current state.The Initial Audit PromptCopy this prompt into ChatGPT, Claude, or your preferred LLM:
What AI Will Ask YouWhen Barry guides executives through this process, AI typically asks questions like:About Creative Work:"What tasks do you find most creative, interesting, and rewarding? List them all out.""When do you feel most energized during your workday?""What work would you do more of if you had unlimited time?"About Routine Work:"What tasks are time-sinks that feel like necessary evils?""What do you find yourself repeatedly doing that could be standardized?""What work drains your energy without adding strategic value?"About Time Allocation:"What percentage do you spend on creative work versus routine work?""Over the last year, how much time have you spent on these categories?" (You can even connect your calendar)Example: The Business Expenses AutomationBarry provides a concrete example: "Say it identifies you spend 5 hours a week doing your business expenses. Necessary, because your accountant wants those things. The machine could say to you, 'Instead of spending those 5 hours manually capturing expenses each month, why don't you try to automate it like this?'"Potential Solution: Use an app where you photograph receipts as they happen, automatically transcribe them into a spreadsheet, and capture images—transforming 90 minutes of weekly work into 30 seconds.The Key Insight: You can ask the tool itself to help you identify where it can help you. This self-teaching capability is revolutionary.Building Your AI workflow Prompt Library: From Novice to ExpertOne of the most powerful aspects of AI executive workflow automation is building a library of prompts that consistently deliver high-quality results. But your prompts should evolve dramatically as you progress.The Progression of Prompt SophisticationNovice Stage (Months 0-2):Barry's reflection: "If I look at prompts I wrote even 6 months ago, I was basically saying 'formulate me a strategy to take over the world' and I was excited by the response it gave me."Novice prompt characteristics:Extremely broad and vagueNo context providedGeneric asks like "make this better"One-sentence requestsSatisfied with any responseExample novice prompts:"Write an email to my team""Summarize this meeting""Give me marketing ideas"Intermediate Stage (Months 3-6):As you gain experience, you start adding context and specificity.Intermediate prompt characteristics:Include relevant contextSpecify desired formatProvide examples of good/badMulti-paragraph instructionsRefine based on initial outputExample intermediate prompts:"Write a follow-up email to my coaching client based on this transcript. Include: key themes discussed, behavior shifts identified, agreed actions, and prep for next session.""Summarize this leadership team meeting. Focus on: decisions made, who owns each action, deadlines, and any concerns raised."Expert Stage (Months 6+):Barry describes his current approach: "Now I'm able to ask questions like, 'I'm thinking about a very specific niche, in this micro-market, in this sub-segment, in this town, in that village'—the specificity that I'm able to get to, which the machine has forced me to do, because I no longer am satisfied with the initial response."Expert prompt characteristics:Extremely specific and detailedInclude multiple data sourcesRequest multiple perspectivesAsk for disconfirming evidenceIterate through multiple roundsBuild on previous conversationsExample expert prompts:"Analyze these 10 meeting transcripts from Q4 2024. Identify: (1) recurring themes about Project Phoenix, (2) implicit assumptions our leadership team is making, (3) perspectives we might be missing from stakeholders not in these meetings, (4) 3 strategic options ranging from conservative to transformational, (5) risks with each option, (6) 10 hard questions a skeptical board member would ask me."Your Prompt Library StructureCreate a document with categorized prompts for common AI executive workflow automation tasks:Meeting Follow-Up Prompts:One-on-one coaching sessionsTeam meetingsClient/stakeholder meetingsStrategy sessionsWeekly business reviewsAnalysis Prompts:Market research synthesisCompetitive analysisPattern identification across meetingsBlind spot detectionRisk assessmentCommunication Prompts:Email drafting for different audiencesPresentation outline creationExecutive summary generationDifficult conversation preparationStrategic Thinking Prompts:Scenario planningDecision-making frameworksInnovation ideationProblem-solving for complex challengesThe Weekly Business Review: Barry's Signature AI Executive Workflow AutomationThis is the workflow that demonstrates the full power of systematic AI executive workflow automation. Barry uses this every Friday at 3:30 PM.The System OverviewTraditional Approach:Time required: 30+ minutesMental energy: HighOutput: Half-page summaryConsistency: Varies based on energy/memoryData sources: Fragmented notes and memoryAI Executive Workflow Automation Approach:Time required: 90 secondsMental energy: Low (review only)Output: Comprehensive multi-page analysisConsistency: Perfect every timeData sources: Complete transcripts of all meetingsHow to Build Your Weekly Business Review SystemStep 1: Capture All Meeting DataEnsure you have AI copilots recording and transcribing every meeting throughout the week. Barry uses Otter.ai, but Fireflies, Fathom, or Grain work equally well.Step 2: Create Your Master PromptHere's a template based on Barry's approach:
Step 3: Establish the RoutineSet a recurring calendar block: Friday 3:30-4:00 PMDownload all week's transcriptsRun the promptReview output for 90 secondsMake any necessary tweaksSend to teamStep 4: Track MetricsBarry tracks several AI executive workflow automation metrics:Completion Metrics:Average time from action creation to closurePercentage of actions completed on timeNumber of actions rolling over week-to-weekEngagement Metrics:Email open rate (are people reading it?)Response rate (are people engaging with it?)Feedback quality (are people finding it valuable?)Efficiency Metrics:Time spent creating the reviewNumber of follow-up questions neededConsistency of deliveryThe Continuous Improvement LoopAs Barry explains: "You start getting all these metrics that help you improve the weekly business review. If only 5% of people are opening it, like a newsletter, you're probably like, 'Well, there's probably something wrong with this content, I need to tweak it a bit.' I can experiment.