AI Productivity Playbook

Foundations of Artificial Intelligence

• What artificial intelligence means in practical terms
  • Narrow AI versus general capability
  • Generative AI versus analytical and rules based systems
• How modern AI systems work conceptually
  • Training data and learned patterns
  • Tokens, context, and why phrasing matters
  • Why AI can sound confident and still be wrong
• Core building blocks explained simply
  • Large language models as prediction engines
  • Embeddings as representations of meaning
  • Retrieval-augmented generation: use source context and still check evidence.
  • Tools and actions as extensions beyond text
• Strengths and limitations
  • Pattern recognition at scale
  • Creativity versus precision
  • Situations where human judgment must lead
• Reliability and trust
  • Hallucinations and fabrication
  • Incomplete context and hidden assumptions
  • Over summarization and loss of nuance
• Professional usage mindset
  • AI as an assistant, not an authority or accountable human colleague.
  • Review, verification, and accountability habits

The AI Landscape and Ecosystem

• Major categories of AI tools
  • Conversational assistants
  • Research and search driven systems
  • Office suite copilots
  • Agentic and task execution systems
  • Creative and generative media tools
  • Automation and workflow platforms
  • Local and command line driven tools
• Where AI appears in daily work
  • Browsers and search
  • Desktop and operating systems
  • Mobile devices
  • Embedded inside enterprise platforms
• Evaluating tools for professional use
  • Accuracy and consistency
  • Transparency and traceability
  • File and data handling capabilities
  • Privacy and compliance considerations
  • Integration with existing workflows
  • Cost models and licensing realities
• The shift toward integrated AI environments
  • AI embedded directly into documents, meetings, and systems
  • Cross-application context only where integrations and permissions support it.
  • AI as a layer rather than a destination

Prompting and Context Design for Productivity

• Prompt structure to test and refine for the chosen model/task
  • Role definition
  • Objective clarity
  • Context and constraints
  • Output format specification
  • Quality criteria and validation checks
• Prompting patterns for everyday work
  • Brainstorming and ideation
  • Drafting structured documents
  • Editing, rewriting, and tone adjustment
  • Summarization for decisions and actions
  • Planning, outlining, and sequencing work
• Iterative collaboration with AI
  • Ask first workflows
  • Critique and revise loops
  • Multiple option generation
  • Comparative evaluation
• Building a personal prompt system
  • Reusable prompt templates
  • Task specific prompt libraries
  • Prompt adaptation for different audiences

Conversational and General Purpose AI Assistants

• Capabilities and use cases of leading assistants
  • Drafting and editing written communication
  • Analyzing documents and reports
  • Supporting decision making
  • Acting as a thinking partner
• ChatGPT and ChatGPT Work, where available
  • Selected text, file, image and voice capabilities subject to account/platform access.
  • Long form drafting and refinement workflows
  • Personal productivity and planning use cases
• Google Gemini
  • Search-grounded assistance where supported; inspect sources and accuracy.
  • Research oriented workflows
  • Deep analysis and exploration patterns
• Microsoft Copilot ecosystem
  • Copilot chat versus application embedded copilots
  • Word, Excel, PowerPoint, Outlook, and Teams usage patterns
  • Organisational grounding subject to licences, configuration and existing access permissions.
• Anthropic Claude
  • Document analysis within supported file/context limits.
  • Structured writing and analysis
• Meta AI assistants
  • Messaging based assistance
  • Lightweight everyday queries
• Choosing the right assistant for the task
  • Creativity versus precision
  • Speed versus depth
  • Context length and document handling

Research, Knowledge, and Document Intelligence Tools

• Research first AI systems
  • Source grounded answers
  • Citation driven exploration
• Research notebooks and knowledge bases
  • Gemini Notebook (formerly NotebookLM): source-grounded workflows where available.
    • Working with trusted sources
    • Asking questions of your own documents
    • Generating briefs, FAQs, and summaries
  • AI inside knowledge management platforms
    • Notion AI
    • Turning notes into structured knowledge
    • Project and decision documentation
• Document processing and understanding
  • Ingesting PDFs, scans, and reports
  • Extracting tables, key facts, and obligations
  • Draft contract/policy summaries and questions for specialist review.
  • Potential compliance/risk issues for verification, not an automated compliance determination.

Agentic AI and Task Execution Systems

• What agentic AI means in practice
  • Planning steps toward a goal
  • Executing across tools
  • Monitoring progress and adjusting
• Manus and selected task-execution tools
  • Multi-step task execution in a bounded environment with reviewed outputs.
  • Research to output pipelines
  • Selected presentation workflows where supported.
  • Multi document synthesis
  • Operational task delegation
• Agent patterns for professional work
  • Research agents
  • Reporting and analysis agents
  • Content production agents
  • Operations and administration agents
• Guardrails for agentic systems
  • Human approval checkpoints
  • Scope limitation
  • Auditability and traceability

Presentation and Content Creation at Scale

• AI assisted presentation workflows
  • Outline to narrative development
  • Slide structure and flow
  • Speaker notes and talking points
• Advanced presentation generation
  • Selected agent-assisted deck creation, such as available Manus workflows.
    • Ingesting briefs, reports, and data
    • Generating structured slide stories
    • Iterative refinement cycles
• Office suite based presentation support
  • Microsoft PowerPoint with Copilot
  • Available AI-assisted Slides workflows; confirm account features.
• Quality control for presentations
  • Message hierarchy
  • Audience alignment
  • Visual density and clarity

Generative AI Beyond Text

• Image generation for professional use
  • Marketing visuals
  • Concept imagery
  • Presentation graphics
• Design and branding tools
  • Canva AI
    • Selected design-generation workflows where available.
    • Brand kit consistency
    • Social and marketing asset generation
• Website-builder demonstrations without coding
  • AI website builders
  • Landing page generation
  • Content structure and copy generation
• Video and audio generation
  • Explainer videos
  • Voiceovers and narration
  • Training and onboarding materials
• Responsible use of generative media
  • Brand alignment
  • Accuracy and claims
  • Consent, authorised assets, usage terms and synthetic-media transparency where relevant.

Automation Platforms and AI Driven Workflows

• Automation fundamentals
  • Triggers, actions, and conditions
  • Human in the loop design
• Microsoft Power Platform
  • Power Automate with AI models
  • Document processing flows
  • Email, forms, and approvals automation
• Available AI model/service integration capabilities
  • Understand the selected platform's supported model management and access controls.
  • Connecting AI services to workflows
• Cross system automation use cases
  • Document intake to summary
  • Meeting notes to task creation
  • Reporting pipelines
• Governance and safety
  • Permissions
  • Error handling
  • Monitoring outcomes

Command Line and Keyboard Driven AI Tools

• Understanding CLI style AI interaction
  • Speed focused workflows
  • Minimal interface environments
• Gemini CLI: optional prepared demonstration
  • Text analysis and transformation
  • Selected non-interactive/batch examples with scoped permissions.
• Local-model environments and their limits
  • Ollama
  • Running models locally
  • Local inference and privacy considerations: local execution alone does not secure all files or integrations.
  • Offline inference only with a suitable downloaded model and a workflow that does not call cloud services.
• When command line tools make sense
  • High volume text processing
  • Sensitive-data requirements: assess every connected service, local environment and policy before use.
  • Power user productivity habits

AI for Finance, Accounting, and Operations

• AI assisted financial workflows
  • Invoice processing
  • Expense categorization
  • Financial summaries and reports
• Agentic accounting assistance
  • Illustrative operational-finance assistance using an available agent, with checks.
  • Cross system data synthesis
  • Audit-preparation drafts, not audit evidence or conclusions.
• Integration with automation platforms
  • Power Automate financial workflows
  • AI driven document extraction
• Risk and accuracy considerations
  • Verify calculations, source data, extracted fields and professional assumptions.
  • Human oversight checkpoints

Building a Personal AI Productivity System

• Designing daily AI workflows
  • Email and communication
  • Meetings and follow ups
  • Document creation
• Weekly and monthly workflows
  • Planning and prioritization
  • Reporting and updates
  • Retrospectives and reviews
• Tool stack design
  • Primary assistant
  • Research and document tools
  • Automation layer
  • Creative generation tools
• Measuring productivity impact
  • Time saved
  • Quality improvements
  • Reduced rework and friction