RC
Rupert Chesman
or Space to navigate
1 / 173
Productivity Pathway

AI Productivity Systems

Build your personal AI productivity system — from first wins to repeatable workflows.

Welcome

Introductions

What does your current workday look like — and where do you feel the friction?

"

Productivity is never an accident. It is always the result of a commitment to excellence, intelligent planning, and focused effort.

Paul J. Meyer

Course Overview

Eight Modules

  • Module 1 — The Productivity Mindset Shift
  • Module 2 — AI Daily Workflows
  • Module 3 — AI Meeting Systems
  • Module 4 — AI Email & Communication
  • Module 5 — AI Research & Analysis
  • Module 6 — AI Note-Taking & Knowledge
  • Module 7 — AI Task & Project Management
  • Module 8 — Building Your Personal AI System

Course Philosophy

  • Systems, not tricks — build workflows that survive model churn
  • Practical and actionable — real prompts, real workflows
  • Model-aware — choose the right tool for the job
  • Low friction — habits that stick without constant effort
  • Progressive — from quick wins to a full personal operating system
Module 1

The Productivity Mindset Shift

From tool-user to system-builder — rethinking how AI fits into your work.

01

The Productivity Mindset Shift

Objectives

What we're
Learning Today

  • Shift from isolated prompts to connected systems
  • Understand the Human Upgrade Loop
  • Design your ideal AI-assisted day
  • Audit your current workflow for AI opportunities
Lesson 1.1

From Tool-User to System-Builder

AI productivity is a system, not a collection of isolated prompts.

"

The amateur uses AI for one-off tasks.

The professional builds systems that compound over time.

The system-builder mindset

Tool-User vs System-Builder

  • Opens ChatGPT when stuck
  • Writes new prompts every time
  • Gets inconsistent results
  • AI feels like extra work
  • Dependent on one model
  • Has reusable prompt libraries
  • Builds workflows that repeat
  • Gets reliable, predictable output
  • AI saves hours every week
  • Chooses the right model for each task

What Makes a Productivity System?

  • Reusable prompts stored and refined over time
  • Defined workflows for recurring tasks
  • Clear handoff points between you and AI
  • Review loops that improve the system continuously
  • Model selection based on task type, not habit

72%

of knowledge workers say they spend more time on process than on actual creative work (McKinsey 2025)

Lesson 1.2

The Human Upgrade Loop

A continuous improvement cycle for human-AI collaboration.

The Human Upgrade Loop

  • Step 1: Identify — Find a repetitive or time-consuming task
  • Step 2: Delegate — Give the task to AI with clear instructions
  • Step 3: Review — Evaluate the output critically
  • Step 4: Refine — Improve your prompt or workflow
  • Step 5: Systematise — Save the refined workflow for reuse
  • Repeat — Each cycle makes the system smarter

Where the Loop Breaks Down

  • Skipping the Review step — accepting output uncritically
  • Never reaching Systematise — reinventing prompts each time
  • Trying to automate judgment-heavy work too early
  • Not adjusting when models update or change behaviour
  • Over-engineering — building a system for a task you do once

Starting Your First Loop

Think about a task you did this week that took longer than it should. Ask yourself: 1. Could I explain this task clearly to a smart assistant? 2. Would the output be roughly the same each time? 3. Do I do this more than once a month? If yes to all three, this is your first upgrade loop candidate.

Lesson 1.3

Designing Your AI Day

Where AI fits into your natural work rhythm.

The AI-Assisted Workday

  • Morning: AI helps plan and prioritise your day
  • Mid-morning: AI handles research and first drafts
  • Lunchtime: AI processes meeting notes from the morning
  • Afternoon: AI assists with communication and follow-ups
  • End of day: AI summarises progress and prepares tomorrow
  • The key: AI handles the process, you handle the judgment

Choosing the Right Model

  • GPT-5.5 — Complex work
  • Deep analysis, strategy documents
  • Multi-step reasoning tasks
  • Creative content with nuance
  • When quality matters most
  • GPT-5.5 Instant — Fast drafting
  • Quick email replies
  • Simple reformatting
  • Brainstorming ideas
  • When speed matters most

Choosing the Right Model (cont.)

  • Claude Opus 4.7 — Reasoning
  • Document analysis
  • Knowledge-heavy work
  • Complex coding tasks
  • Long-form writing and editing
  • Gemini 3.1 Pro — Grounded workflows
  • Multi-step research tasks
  • Google ecosystem integration
  • Data analysis with sources
  • Tasks needing current information
Lesson 1.4

Audit Your Current Workflow

Before you build, understand where you are now.

The Three Task Categories

  • Repetitive tasks — Same steps, same output, every time
  • Judgment-heavy tasks — Require your expertise and decisions
  • Communication-heavy tasks — Drafting, replying, summarising
  • Most people discover 40-60% of their work is repetitive
  • AI handles repetitive best, assists with judgment and communication
Activity

Personal Workflow Audit

15 min

List your top 10 weekly tasks. For each one, mark it: [R] Repetitive — Same steps each time [J] Judgment-heavy — Needs your expertise [C] Communication-heavy — Writing/replying Now circle the top 3 that consume the most time. These are your first AI system candidates.

4.1 hours

per day spent on tasks that could be partially or fully automated with AI (Accenture 2025 Workforce Study)

Module 2

AI Daily Workflows

Build an AI-assisted workday rhythm that survives model churn.

02

AI Daily Workflows

Recap

What we
Learnt

  • Shifted from tool-user to system-builder mindset
  • Learned the Human Upgrade Loop
  • Mapped models to task types
  • Audited personal workflow for AI opportunities
Objectives

What we're
Learning Today

  • Set up an AI morning routine
  • Build daily planning and prioritisation prompts
  • Create end-of-day review workflows
  • Design weekly and monthly AI reviews
Lesson 2.1

AI Morning Routine

Start every day with clarity, not chaos.

The 10-Minute AI Morning

  • Step 1: Open your AI tool of choice (2 min)
  • Step 2: Paste your calendar + task list for today (1 min)
  • Step 3: Ask AI to identify your top 3 priorities (2 min)
  • Step 4: Ask for a realistic time-blocked schedule (3 min)
  • Step 5: Review and adjust — you make the final call (2 min)
  • Total: 10 minutes for a focused, intentional day

Morning Planning Prompt

Act as my executive assistant with deep understanding of productivity and energy management. Here is my calendar for today: [Paste calendar] Here are my outstanding tasks: [Paste task list] Please: 1. Identify my top 3 highest-impact tasks for today 2. Create a time-blocked schedule that groups similar work 3. Flag any meetings I should prepare for 4. Suggest one task I should defer or delegate

"

The morning routine is not about AI doing your thinking.

It is about AI clearing the noise so you can think better.

Systems thinking in practice

Lesson 2.2

Daily Planning & Prioritisation

Let AI help you decide what matters most.

Why AI-Assisted Prioritisation Works

  • You already know what needs doing — AI helps you sort it
  • AI removes recency bias — your latest email is not always urgent
  • AI can cross-reference deadlines, dependencies, and energy
  • The goal is not autopilot — it is faster, better-informed decisions
  • You always make the final call

Building a Reusable Planning Template

  • Create a standard format for your daily input (calendar + tasks)
  • Save your best planning prompt as a template
  • Iterate the template weekly — refine based on what works
  • Store templates where you can find them (notes app, doc, custom instruction)
  • A good template should work with any model
Lesson 2.3

Prioritisation Systems with AI

Classic frameworks, supercharged with AI assistance.

Eisenhower Matrix with AI

  • Urgent + Important — Do now, AI can help you prepare
  • Important + Not Urgent — Schedule, AI can help you plan
  • Urgent + Not Important — Delegate, AI can draft the delegation
  • Not Urgent + Not Important — Eliminate or defer
  • AI is excellent at sorting tasks into these quadrants
  • You provide the judgment on what is truly important

Eisenhower Matrix Prompt

Act as a productivity coach who uses the Eisenhower Matrix. Here are my tasks for this week: [Paste task list] Please sort each task into the Eisenhower Matrix: - Urgent & Important (Do First) - Important, Not Urgent (Schedule) - Urgent, Not Important (Delegate) - Neither (Eliminate) For each task, explain your reasoning in one sentence.

Energy Mapping

  • Match task difficulty to your energy levels throughout the day
  • Morning energy peak — deep work, strategy, complex writing
  • Post-lunch dip — admin, emails, routine tasks
  • Afternoon recovery — meetings, collaborative work
  • AI can help you build an energy-aware schedule
  • Track your energy for one week to find your pattern
Activity

Build Your Priority System

15 min

Take your task list for this week and: 1. Use the Eisenhower Matrix prompt to sort your tasks 2. Map your typical energy levels across the day 3. Create a time-blocked schedule matching high-energy tasks to peak hours Save the prompt template for reuse.

Lesson 2.4

End-of-Day Review

Close each day with intention and set up tomorrow for success.

The 5-Minute AI Review

  • What did I accomplish today? (list completions)
  • What did I not finish? (identify carry-overs)
  • What surprised me? (capture unexpected tasks)
  • What should I start with tomorrow?
  • AI turns this into a structured daily log
  • Over time, you build a record of patterns and progress

End-of-Day Review Prompt

Act as my accountability partner. Here is what I planned for today: [Paste morning plan] Here is what actually happened: [Brief summary of your day] Please: 1. Compare planned vs actual — what got done and what slipped 2. Identify any patterns (meetings running over, tasks taking longer) 3. Suggest my top 3 priorities for tomorrow 4. Note any follow-ups I need to send

Lesson 2.5

Weekly & Monthly Reviews

Zoom out regularly to improve the system itself.

Weekly Review with AI

  • Every Friday: Feed AI your daily logs from the week
  • Ask for patterns — what tasks keep slipping?
  • Identify time thieves — where did unplanned work come from?
  • Celebrate wins — what went well this week?
  • Plan next week with lessons learned
  • Takes 15-20 minutes, saves hours of drift

Monthly System Review

  • Review your prompt templates — are they still working?
  • Check your model choices — has anything improved?
  • Assess your workflows — any new bottlenecks?
  • Update your custom instructions if your role has changed
  • Set one improvement goal for next month
  • The system that does not evolve becomes the bottleneck
Activity

Design Your Review Rhythm

15 min

Create your personal review system: 1. Write a daily review prompt you will actually use (keep it under 5 minutes) 2. Write a weekly review prompt (Friday, 15 minutes) 3. Set a monthly calendar reminder for a system audit Test your daily review prompt right now with today's work.

23 minutes

average time to refocus after an interruption — structured reviews reduce this by batching reflection (University of California, Irvine)

Module 3

AI Meeting Systems

Turn meetings into preparation, capture, action, and accountability systems.

03

AI Meeting Systems

Recap

What we
Learnt

  • Built an AI morning routine
  • Created daily planning and prioritisation templates
  • Designed end-of-day review workflows
  • Set up weekly and monthly review rhythms
Objectives

What we're
Learning Today

  • Create AI-powered meeting preparation workflows
  • Build agendas with AI assistance
  • Set up AI note-taking systems
  • Automate action item extraction and follow-ups
Lesson 3.1

Meeting Preparation Workflows

Never walk into a meeting unprepared again.

The Preparation System

  • Step 1: Feed AI the meeting invite, attendee list, and context
  • Step 2: Ask for background on attendees and topics
  • Step 3: Generate talking points and potential questions
  • Step 4: Identify decisions that need to be made
  • Step 5: Draft any pre-read materials needed
  • Total prep time: 5-10 minutes vs 30 minutes manually

Meeting Prep Prompt

Act as my executive briefing assistant. I have a meeting in 2 hours: - Topic: [Meeting subject] - Attendees: [List names and roles] - Context: [What this meeting is about] - Previous meeting notes: [Paste if available] Please prepare: 1. A 3-sentence briefing on the key context I need 2. Three questions I should be ready to answer 3. Two questions I should ask 4. Any decisions that likely need to be made

Lesson 3.2

Agenda Creation with AI

Structured agendas that keep meetings on track.

AI-Generated Agendas

  • Give AI the meeting purpose, attendees, and time limit
  • AI creates time-boxed agenda items with owners
  • Include decision points and expected outcomes
  • Add a parking lot section for off-topic items
  • Send the agenda 24 hours in advance
  • The best meeting is the one that did not need to happen

Agenda Builder Prompt

Act as a meeting facilitator. Create an agenda for a [duration]-minute meeting: - Purpose: [What we need to accomplish] - Attendees: [Names and roles] - Key decisions needed: [List any] Please create a time-boxed agenda with: 1. Each item with a time allocation and owner 2. A clear expected outcome for each item 3. A 5-minute buffer for wrap-up and action items 4. A parking lot section

Lesson 3.3

AI Note-Taking Systems

Capture everything without losing focus on the conversation.

Platform-Native vs Model-Assisted

  • Teams Copilot
  • Built into Microsoft Teams
  • Real-time transcription
  • Auto-generated summaries
  • Action item detection
  • Requires Microsoft 365 licence
  • Google Meet Notes
  • Built into Google Meet
  • Automatic note-taking
  • Integrates with Google Docs
  • Searchable transcript
  • Requires Google Workspace

Post-Meeting Processing

  • If no native tool: record and transcribe with Otter.ai or similar
  • Paste transcript into Claude Opus 4.7 for deep analysis
  • Ask for: summary, action items, decisions, open questions
  • Claude excels at extracting structured data from messy transcripts
  • Store processed notes in a consistent location
  • Privacy note: check your organisation's recording policies
Lesson 3.4

Action Item Extraction

Turn meeting notes into accountable to-do items.

Action Item Extraction Prompt

Act as a project coordinator. Here are the notes from today's meeting: [Paste meeting notes or transcript] Please extract: 1. All action items with the person responsible 2. All decisions that were made 3. All open questions that need follow-up 4. Any deadlines mentioned Format each action item as: [Owner] — [Task] — [Due date if mentioned]

Making Action Items Stick

  • Assign a single owner to every action item
  • Include a deadline — even if approximate
  • Send the action items within 1 hour of the meeting
  • Follow up at the next meeting or check-in
  • AI can draft the follow-up email for you
  • No owner + no deadline = no action
Lesson 3.5

Follow-Up Automation

Close the loop without adding to your to-do list.

The Follow-Up Workflow

  • AI extracts action items immediately after the meeting
  • AI drafts a summary email to all attendees
  • AI creates calendar reminders for key deadlines
  • AI drafts check-in messages for 48 hours later
  • You review and send — AI handles the process
  • The follow-up is where most meeting value is lost
Activity

Build Your Meeting System

15 min

Think about your most recent meeting: 1. Use the action item extraction prompt on your notes 2. Draft a follow-up email using AI 3. Create a meeting prep template you can reuse Save all three prompts in your template library.

$37 billion

lost annually to unproductive meetings in the US alone (Harvard Business Review)

Module 4

AI Email & Communication

Tone-stable, role-aware communication systems that save hours.

04

AI Email & Communication

Recap

What we
Learnt

  • Built meeting preparation workflows
  • Created AI-powered agendas and note-taking
  • Extracted action items with AI
  • Designed follow-up automation
Objectives

What we're
Learning Today

  • Build email drafting frameworks and templates
  • Master tone adjustment and voice matching
  • Create inbox management strategies
  • Design template systems and response chains
Lesson 4.1

Drafting Frameworks & Templates

Never stare at a blank email again.

The Email Drafting System

  • Define your standard email types (update, request, reply, escalation)
  • Create a template prompt for each type
  • Include your voice and tone preferences in each template
  • Use GPT-5.5 Instant for routine emails, GPT-5.5 for sensitive ones
  • Build a library of 8-10 templates that cover 90% of your emails
  • Refine templates monthly based on what works

Email Draft Prompt

Act as a professional communication specialist who writes in a warm but direct style. I need to write an email: - To: [Recipient and relationship] - Purpose: [What I need from them] - Context: [Background they need] - Tone: [Professional/casual/urgent/diplomatic] - Length: [Brief/medium/detailed] Please draft the email. Keep it scannable with short paragraphs. End with a clear call to action and a specific deadline.

Email Types to Template

  • Status update — Progress report to stakeholders
  • Request — Asking someone to do something
  • Reply — Responding to a question or request
  • Escalation — Raising an issue that needs attention
  • Introduction — Connecting two people
  • Follow-up — Checking in on a previous conversation
  • Decline — Saying no with grace
  • Thank you — Acknowledging help or contribution
Lesson 4.2

Tone Adjustment & Voice Matching

Sound like yourself, even when AI does the writing.

Voice Matching Techniques

  • Give AI examples of your actual writing to learn your style
  • Specify tone modifiers: "warm but professional", "direct but kind"
  • Ask AI to adjust formality level on a scale of 1-10
  • Use custom instructions to embed your voice permanently
  • Always read AI drafts aloud — does it sound like you?
  • If it sounds generic, add more specific constraints

When to Use Fast vs Capable Models

  • GPT-5.5 Instant / Fast models
  • Quick replies to simple questions
  • Internal team messages
  • Routine status updates
  • Formatting and proofreading
  • When speed matters most
  • GPT-5.5 / Claude Opus 4.7
  • Sensitive or high-stakes emails
  • Client-facing communication
  • Conflict resolution messages
  • Executive-level updates
  • When nuance matters most

Tone Adjustment Prompt

Here is an email I need to send: [Paste your draft] Please rewrite this email with the following adjustments: - Tone: [More formal / more casual / more diplomatic / more direct] - Length: [Shorter / same / more detailed] - Emphasis: [What point should come through strongest] Keep my core meaning but adjust the delivery. Make it sound natural, not robotic.

Lesson 4.3

Inbox Management Strategies

Process email in batches, not in real time.

The Batch Processing Approach

  • Check email 3 times per day, not constantly
  • Morning batch: Triage and quick replies (15 min)
  • Midday batch: Respond to items needing thought (20 min)
  • End of day: Follow-ups and tomorrow prep (10 min)
  • Use AI to draft replies during each batch
  • The goal: inbox as a system, not an interruption

AI-Assisted Triage

  • Copy a batch of email subjects and senders into AI
  • Ask AI to categorise: Urgent, Important, FYI, Delegate
  • AI can draft quick replies for the simple ones
  • You focus your energy on the ones that need real thought
  • Over time, AI learns your patterns from custom instructions
  • Never process emails one at a time when batching is faster
Lesson 4.4

Template Systems & Response Chains

Reusable response chains for predictable communication.

Building Response Chains

  • A response chain is a series of templates for a common sequence
  • Example: New client inquiry → Acknowledgment → Proposal → Follow-up
  • Example: Complaint → Acknowledgment → Investigation → Resolution
  • Each step has a template prompt that takes context from the previous
  • Store chains in a document or notes app for quick access
  • Update chains quarterly based on what you are actually sending
Activity

Build Your Email System

15 min

Create your personal email toolkit: 1. Write template prompts for your 3 most common email types 2. Create one complete response chain (3-4 emails in sequence) 3. Set your voice/tone preferences as a reusable instruction Test one template with a real email you need to send today.

Lesson 4.5

Communication Workflow Automation

Connecting your communication system to your broader workflow.

Automation Opportunities

  • Meeting follow-up emails sent automatically after note processing
  • Weekly status update emails drafted from your daily logs
  • Client check-in emails triggered by calendar reminders
  • Proposal follow-ups sent on a schedule
  • Internal announcements drafted from project updates
  • The less typing you do, the more thinking you can do

28%

of the average workweek is spent managing email (McKinsey Global Institute)

Module 5

AI Research & Analysis

Search, compare, and synthesise without surrendering your judgment.

05

AI Research & Analysis

Recap

What we
Learnt

  • Built email drafting frameworks and templates
  • Mastered tone adjustment and voice matching
  • Created inbox management strategies
  • Designed template systems and response chains
Objectives

What we're
Learning Today

  • Conduct deep research workflows with AI
  • Evaluate and verify AI-provided sources
  • Synthesise information into decision-ready briefs
  • Build competitive analysis frameworks
Lesson 5.1

Deep Research Workflows

From question to comprehensive answer, systematically.

The Research Stack

  • Layer 1: Quick query — GPT-5.5 Instant for surface-level answers
  • Layer 2: Deep research — Gemini 3.1 Pro for grounded, sourced research
  • Layer 3: Analysis — Claude Opus 4.7 for synthesising multiple sources
  • Layer 4: Verification — Cross-reference across models and sources
  • Use the right layer for the depth you need
  • Do not use Layer 1 tools for Layer 3 problems

Structuring a Research Task

  • Define the question precisely — ambiguity produces vague results
  • Specify the scope — time period, geography, industry
  • Request sources and citations — always
  • Ask for confidence levels on key claims
  • Break large research into sequential sub-questions
  • Provide any documents you already have as context

Deep Research Prompt

Act as a senior research analyst. I need to understand: [Your research question] Scope: - Time period: [e.g., last 12 months] - Geography: [e.g., Australia and New Zealand] - Industry: [e.g., financial services] Please: 1. Provide a comprehensive overview with key findings 2. Cite specific sources for every major claim 3. Mark your confidence level (High/Medium/Low) for each finding 4. Identify gaps where reliable data is unavailable 5. Suggest 3 follow-up questions worth investigating

Lesson 5.2

Source Evaluation & Verification

Trust but verify — every time.

The Verification Framework

  • Step 1: Ask AI for the source of every claim
  • Step 2: Check if the cited source actually exists
  • Step 3: Verify the claim says what AI claims it says
  • Step 4: Cross-reference with a second model or source
  • Step 5: Mark unverifiable claims clearly in your output
  • AI hallucinations are most dangerous when they sound authoritative
"

The most dangerous AI output is the one that is 95% correct.

The 5% that is wrong will be the part you needed most.

On verification discipline

Red Flags in AI Research

  • Very specific statistics without a named source
  • Citations to papers or reports you cannot find
  • Claims that are suspiciously perfectly aligned with your question
  • Overly confident language on nuanced or contested topics
  • Identical phrasing across multiple "different sources"
  • When in doubt: ask the AI "Are you certain? Can you provide the exact URL?"
Lesson 5.3

Synthesis & Report Generation

Turn raw research into decision-ready briefs.

The Synthesis Process

  • Gather research outputs from multiple queries and models
  • Feed all raw material into Claude Opus 4.7 for synthesis
  • Ask for a structured brief: summary, key findings, recommendations
  • Request an executive summary for stakeholders
  • Include a methodology section noting AI tools used
  • Always include limitations and caveats

Synthesis Prompt

Act as a senior analyst preparing a decision brief for leadership. Here is my research from multiple sources: [Paste all research findings] Please create a structured brief: 1. Executive summary (3 sentences max) 2. Key findings (5-7 bullet points with evidence) 3. Risks and uncertainties 4. Recommended actions (prioritised) 5. Appendix: sources used and confidence assessment Write for a time-poor executive who needs to make a decision.

Lesson 5.4

Competitive Analysis Frameworks

Structured competitor intelligence with AI.

AI-Assisted Competitive Analysis

  • Define your competitive set — direct, indirect, aspirational
  • Use Gemini 3.1 Pro for grounded web research on competitors
  • Use Claude Opus 4.7 to analyse competitor documents
  • Create a standard comparison framework (pricing, features, positioning)
  • Update quarterly — set a calendar reminder
  • AI finds the information; you provide the strategic interpretation

Competitive Analysis Template

  • Company overview and positioning
  • Product/service comparison matrix
  • Pricing strategy analysis
  • Content and marketing approach
  • Strengths, weaknesses, opportunities, threats
  • Strategic implications for your business
Lesson 5.5

Research Project Workshop

Apply everything to a real research challenge.

Activity

Research Project

20 min

Choose a real research question from your work: 1. Use the Deep Research Prompt to gather initial findings 2. Verify at least 2 key claims using the Verification Framework 3. Synthesise your findings into a decision brief 4. Share your brief with the group Use at least 2 different AI models in your research.

5x

faster to produce a first draft research brief with AI assistance vs manual research (Deloitte AI Productivity Report 2025)

Module 6

AI Note-Taking & Knowledge

Build a practical second brain — capture, surface, retrieve, connect.

06

AI Note-Taking & Knowledge

Recap

What we
Learnt

  • Conducted deep research workflows with AI
  • Learned to evaluate and verify sources
  • Created synthesis and report generation workflows
  • Built competitive analysis frameworks
Objectives

What we're
Learning Today

  • Design a second brain system with AI
  • Build knowledge capture and retrieval workflows
  • Connect ideas across projects using AI
  • Create a sustainable knowledge maintenance practice
Lesson 6.1

Second Brain Systems with AI

A practical knowledge system that does not require a PhD in organisation.

What Is a Second Brain?

  • A system for capturing, organising, and retrieving knowledge
  • Not about storing everything — about finding what matters
  • Traditional: Notion, Obsidian, Evernote
  • AI-enhanced: AI helps you capture, tag, connect, and surface
  • The best system is the one you actually use
  • Start simple, add complexity only when needed

Traditional vs AI-Enhanced

  • Manual tagging and categorisation
  • You remember where you put things
  • Search by keyword only
  • Connections you notice yourself
  • Maintenance is a chore
  • AI suggests tags and categories
  • AI retrieves based on context
  • Search by meaning and concept
  • AI surfaces unexpected connections
  • AI assists with maintenance

The Minimum Viable Second Brain

  • One capture tool (notes app, document, whatever you already use)
  • One AI tool for processing (Claude Opus 4.7 recommended)
  • A simple folder structure: Projects, Areas, Resources, Archive
  • A weekly 15-minute review to process your capture inbox
  • That is it — do not over-engineer on day one
Lesson 6.2

Knowledge Capture & Retrieval

Capture fast, retrieve faster.

Capture Workflows

  • Meeting notes → AI processes → stored with tags
  • Articles and reports → AI summarises → key takeaways saved
  • Ideas and observations → quick capture → AI categorises later
  • Emails with useful info → forwarded to capture inbox
  • The rule: capture in under 30 seconds or you will not do it
  • Process captured items weekly, not in real time

Knowledge Processing Prompt

Act as my knowledge management assistant. Here is a [meeting note / article / document] I want to store: [Paste content] Please: 1. Write a 2-3 sentence summary 2. Extract the key insights (max 5) 3. Suggest 3-5 tags for categorisation 4. Identify any action items or follow-ups 5. Note any connections to common work topics: [list your key projects/areas]

Lesson 6.3

Connecting Ideas Across Projects

The real value of a knowledge system is in the connections.

AI as Connection Engine

  • Feed AI notes from different projects and ask for connections
  • AI can spot patterns you miss because you are too close to the work
  • Ask: "What themes appear across these three project notes?"
  • Ask: "How does finding X relate to challenge Y?"
  • Schedule monthly connection sessions — 15 minutes of cross-pollination
  • This is where insight lives — at the intersection of ideas

Connection Finding Prompt

Act as a creative thinking partner. Here are notes from three different projects I am working on: Project A: [Paste key notes] Project B: [Paste key notes] Project C: [Paste key notes] Please: 1. Identify any themes or patterns that appear across projects 2. Suggest unexpected connections between ideas 3. Flag any contradictions or tensions worth exploring 4. Propose one insight that combines learning from multiple projects

Lesson 6.4

Zettelkasten with AI

The atomic note method, supercharged.

AI-Enhanced Zettelkasten

  • Zettelkasten: one idea per note, notes linked by connection
  • AI helps break long content into atomic notes
  • AI suggests links between new notes and existing ones
  • AI helps you write "link notes" explaining why two ideas connect
  • Tools: Obsidian + AI plugin, Notion AI, or manual with Claude
  • Warning: do not build a Zettelkasten unless you enjoy the process
Lesson 6.5

Building Your Knowledge Base

A living library of everything you have learned.

Knowledge Base Maintenance

  • Archive notes you have not accessed in 6 months
  • Update outdated information when you encounter it
  • Merge duplicate or overlapping notes
  • AI can help identify stale or redundant entries
  • The goal is a lean, useful system — not a comprehensive archive
  • A knowledge base you cannot search is worse than no knowledge base
Activity

Start Your Second Brain

15 min

Right now: 1. Choose your capture tool (notes app, doc, etc.) 2. Create four folders: Projects, Areas, Resources, Archive 3. Process one piece of content using the Knowledge Processing Prompt 4. Set a weekly 15-minute review reminder Keep it simple — you can always add complexity later.

"

Your second brain should make you smarter, not busier.

If maintaining it feels like a second job, simplify.

On knowledge systems

Module 7

AI Task & Project Management

Connect AI to your delivery work — using the tools you already have.

07

AI Task & Project Management

Recap

What we
Learnt

  • Designed a second brain system with AI
  • Built knowledge capture and retrieval workflows
  • Connected ideas across projects
  • Created a knowledge base maintenance practice
Objectives

What we're
Learning Today

  • Integrate AI with your existing task management tools
  • Use AI for project planning and breakdown
  • Automate status updates and reporting
  • Set up deadline tracking and accountability
Lesson 7.1

Integration with Existing Tools

AI should enhance your tools, not replace them.

The Integration Principle

  • Do not switch tools just to use AI — integrate AI into what you have
  • Trello, Asana, Monday, Jira — all work with AI-assisted workflows
  • Three levels of AI integration:
  • 1. Manual support — copy-paste between AI and your tool
  • 2. Rule-based automation — Zapier/Make connects AI to your tool
  • 3. Agentic orchestration — AI operates your tool directly
  • Start at Level 1, move up only when the ROI is clear

Three Levels of Integration

  • Level 1: Manual Support
  • Copy task list to AI
  • AI helps prioritise
  • You update your tool
  • Zero setup required
  • Good for: getting started
  • Level 2: Rule-Based Automation
  • Zapier connects AI to your tool
  • Automatic task creation from emails
  • AI-generated status updates
  • Some setup required
  • Good for: recurring workflows

Level 3: Agentic Orchestration

  • AI directly reads and updates your project tool
  • Creates tasks, assigns owners, sets deadlines
  • Monitors progress and flags risks
  • Generates reports without you asking
  • Requires careful permission management
  • Good for: teams with mature AI practices
Lesson 7.2

Project Planning with AI

From vague scope to detailed work breakdown in minutes.

Project Breakdown Prompt

Act as a senior project manager. I am starting a new project: - Project: [Name and description] - Deadline: [Target completion date] - Team: [Who is available and their roles] - Constraints: [Budget, dependencies, risks] Please create: 1. A work breakdown structure with all major phases 2. Key milestones with target dates 3. Dependencies between tasks 4. Risk register with mitigation strategies 5. A suggested weekly check-in cadence

AI Planning Best Practices

  • Use AI for the first draft — then refine with your team
  • Ask AI to challenge your timeline: "What could go wrong?"
  • Have AI identify hidden dependencies you might miss
  • Use AI to estimate task durations based on similar projects
  • Always validate AI estimates with experienced team members
  • AI produces the structure; your experience provides the reality check
Lesson 7.3

Status Updates & Reporting

Never write a status report from scratch again.

AI-Generated Status Updates

  • Feed AI your task completions, blockers, and next steps
  • AI drafts a status update tailored to your audience
  • Executive version: 3-sentence summary with RAG status
  • Team version: detailed progress with action items
  • Create templates for weekly, monthly, and ad-hoc updates
  • The update that gets sent consistently is better than the perfect one

Status Update Prompt

Act as my project coordinator. Here is my project status: - Completed this week: [List items] - In progress: [List items with % complete] - Blocked: [List items and why] - Upcoming: [Next week priorities] Please create two versions: 1. Executive summary (3 sentences, RAG status, key risks) 2. Team update (detailed, with action items and owners)

Lesson 7.4

Deadline Tracking & Accountability

AI as your accountability partner.

Accountability Systems

  • Weekly AI review of your task list vs deadlines
  • AI flags tasks at risk of slipping before they slip
  • AI drafts check-in messages for delegated tasks
  • Set up a Friday ritual: feed AI your full task list
  • AI identifies the one thing most likely to derail next week
  • Accountability is not about pressure — it is about visibility
Activity

Build Your Project AI Toolkit

15 min

Using your current project management tool: 1. Draft a project breakdown for a real upcoming project 2. Create a status update template you can reuse weekly 3. Set up a Friday accountability review prompt Test the status update template with this week's actual progress.

45%

of project managers say status reporting is their most time-consuming admin task (PMI Pulse of the Profession 2025)

Module 8

Building Your Personal AI System

Combine everything into a durable personal operating system.

08

Building Your Personal AI System

Recap

What we
Learnt

  • Integrated AI with existing task management tools
  • Used AI for project planning and breakdown
  • Automated status updates and reporting
  • Set up deadline tracking and accountability
Objectives

What we're
Learning Today

  • Combine all modules into one coherent system
  • Set up custom instructions and preferences
  • Design workflow automation strategies
  • Build continuous improvement loops
Lesson 8.1

Your Personal AI Operating System

One system, not eight separate tools.

The Personal AI OS

  • Morning: AI planning and prioritisation (Module 2)
  • Meetings: Prep, capture, action items, follow-up (Module 3)
  • Communication: Templates, tone matching, batch processing (Module 4)
  • Research: Structured queries, verification, synthesis (Module 5)
  • Knowledge: Capture, process, connect, maintain (Module 6)
  • Projects: Planning, tracking, reporting, accountability (Module 7)
  • Everything connected by consistent habits and reusable prompts

The Connection Points

  • Morning plan feeds into meeting prep
  • Meeting notes feed into action items and knowledge base
  • Research feeds into communication and project planning
  • End-of-day review feeds into tomorrow's morning plan
  • Weekly review feeds into monthly system improvement
  • Each module outputs become the next module's inputs
"

A system is not a collection of tools.

A system is the connections between them.

On building your AI OS

Lesson 8.2

Custom Instructions & Preferences

Teach AI who you are once, not every conversation.

Setting Up Custom Instructions

  • Your role, industry, and key responsibilities
  • Your communication style and tone preferences
  • Your standard output formats (bullet points, tables, briefs)
  • Common tools and platforms you use
  • Recurring context (team size, company stage, key projects)
  • Update quarterly or when your role changes

Custom Instructions Template

About me: - Role: [Your job title and responsibilities] - Industry: [Your industry and specialisation] - Team: [Who you work with and your reporting structure] - Tools: [Key software and platforms you use daily] Communication preferences: - Style: [Direct/conversational/formal] - Length: [Concise/detailed/varies by context] - Format: [Bullet points/paragraphs/tables] Context I frequently reference: - [Key project 1] - [Key project 2] - [Recurring tasks or responsibilities]

Model-Specific Instructions

  • ChatGPT: Custom Instructions in Settings
  • Claude: Project instructions and system prompts
  • Gemini: Gems and saved preferences
  • Each model stores preferences differently
  • Maintain a master document and adapt for each platform
  • Your custom instructions are a competitive advantage — invest in them
Lesson 8.3

Workflow Automation Strategies

From manual prompts to automated pipelines.

The Automation Ladder

  • Rung 1: Reusable prompts saved in a document
  • Rung 2: Custom GPTs or Claude Projects for specific workflows
  • Rung 3: Zapier Agents connecting AI to your tools
  • Rung 4: Make AI Agents for complex multi-step workflows
  • Rung 5: N8N for self-hosted, fully custom automation
  • Climb one rung at a time — each rung should be stable before moving up

Automation Tools Compared

  • Zapier Agents
  • Easiest to set up
  • Largest app library
  • AI-powered agent mode
  • Best for: simple connections
  • Pricing: per-task billing
  • Make AI Agents
  • Visual workflow builder
  • More complex logic
  • AI steps built in
  • Best for: multi-step workflows
  • Pricing: operations-based

N8N: The Power User Option

  • Open-source workflow automation
  • Self-hosted — your data stays with you
  • Connect any API, including AI models
  • Most flexible but requires more technical setup
  • Community templates for common workflows
  • Best for: teams with technical capability who want full control

Automation Candidates

  • Email → AI triage → Draft response → Your approval → Send
  • Meeting transcript → AI summary → Action items → Task manager
  • Weekly data → AI analysis → Status report → Stakeholder email
  • New client inquiry → AI research → Briefing document → CRM
  • Social mention → AI sentiment analysis → Alert if negative
  • The best automation removes a manual step you already do
Lesson 8.4

Continuous Improvement Loops

The system that does not evolve becomes the bottleneck.

Building Review Loops

  • Daily: Did my prompts work today? Quick mental note
  • Weekly: Which templates need updating? 5-minute check
  • Monthly: Is my system serving me or am I serving it? 30-minute audit
  • Quarterly: Full system review — models, tools, workflows, instructions
  • Annual: Strategic review — is AI changing my role? How should I adapt?
  • Each loop feeds improvements into the next cycle

Measuring ROI

  • Track time saved per week (even rough estimates help)
  • Track quality improvements (fewer revisions, faster approvals)
  • Track consistency (fewer missed follow-ups, better handoffs)
  • Compare before-and-after for specific workflows
  • Share wins with your team to build adoption
  • If you cannot measure improvement, simplify the system

Approval Points

  • Never fully automate without review checkpoints
  • AI drafts, you approve — this is the foundation
  • As trust builds, you can widen the approval window
  • But always maintain a human checkpoint for:
  • — External communication
  • — Financial decisions
  • — Anything with legal implications
  • — Novel situations outside your templates
Lesson 8.5

Your Personalised AI Operating System

From individual practice to team adoption.

Scaling from Individual to Team

  • Start with yourself — prove the system works
  • Document your best workflows in plain language
  • Share one prompt template with a colleague — see if it works for them
  • Create a team prompt library with your most effective templates
  • Run a 30-minute team workshop showing your system
  • Adoption spreads through demonstration, not mandate

Your AI System Blueprint

  • 1. Morning routine prompt (Module 2)
  • 2. Meeting prep template (Module 3)
  • 3. Email drafting templates (Module 4)
  • 4. Research workflow (Module 5)
  • 5. Knowledge capture process (Module 6)
  • 6. Project planning template (Module 7)
  • 7. Custom instructions for each AI model (Module 8)
  • 8. Weekly review checklist (Module 2)
  • Print this list. Check off each one as you build it.
Activity

Build Your Blueprint

20 min

This is your final activity. Create your personal AI System Blueprint: 1. List the top 5 workflows you will implement first (from any module) 2. Set up custom instructions in your primary AI tool right now 3. Create your weekly review prompt and set a calendar reminder 4. Identify one workflow you will automate in the next 30 days 5. Write down your 30-day, 90-day, and 6-month AI goals Share your blueprint with the group.

"

The goal is not to be productive with AI.

The goal is to build a system that makes productivity automatic.

Your AI operating system

Resources

Your Toolkit

Everything you need to get started.

Quick Start Checklist

  • Set up custom instructions in your primary AI tool
  • Save your top 5 prompt templates somewhere accessible
  • Create your morning planning and end-of-day review prompts
  • Set a weekly Friday review reminder
  • Choose one workflow to automate this month
  • Share Act-Explain-Please with one colleague
  • Visit www.rupertchesman.com for additional resources

Recommended Models by Task

  • Complex analysis and strategy — GPT-5.5
  • Quick drafts and simple tasks — GPT-5.5 Instant
  • Reasoning, coding, document analysis — Claude Opus 4.7
  • Grounded research with sources — Gemini 3.1 Pro
  • Each model has strengths — match the model to the task
  • Your custom instructions should be set up in at least two models

Systems, not tricks.

The difference between using AI and being productive with AI.

Q&A

Questions

What would you like to know more about?

Thank You

Thank You

www.rupertchesman.com · © Rupert Chesman 2026