A 12-minute orientation across the AI-for-CX chapter — the four-layer operational ladder, the ethical anchor on deflection, and where to begin

Build the AI triage layer that sits at the front of every CX queue — taxonomy, confidence scoring, and priority stacking

AI-assisted CX replies that sound like your brand — tone presets, de-escalation, and the human-review matrix

Build a knowledge base from real tickets, write findable articles, and maintain them on a cadence

Turn thousands of conversations into themes — sentiment taxonomy, churn-phrase watchlists, and multi-stakeholder distribution

The ethical anchor — name the trap high deflection hides, separate good deflection from relationship-burning, and give leaders push-back language

Build three recursive CX tools — self-updating FAQ, self-learning template factory, and continuous sentiment monitor

A 10-minute orientation across the five-chapter beginner path — what agents are, how they work, where you already use them, when they fail, and how to work alongside them

The mental model shift from AI that answers questions to AI that does work — explained through everyday workplace scenarios, no code required

The three ingredients that turn a chatbot into an agent — tools it can use, memory it can draw on, and goals it can pursue — explained through workplace analogies

A tour of AI agents you probably already encounter at work — Microsoft Copilot, customer service bots, scheduling assistants, and more — with the exact moment each one stops being a chatbot

An honest look at agent limitations — hallucination chains, confident mistakes, and loops — framed as normal challenges with practical responses, not reasons to panic

Practical guidance for working with AI agents daily — writing good instructions, knowing when to intervene, reviewing output, and building your personal agent playbook

Your map to the 7 marketing-specific AI capabilities — pick the right tool per job, extract brand voice, plan campaigns, master Generative Engine Optimization (GEO), build small automations, recognize when AI is your partner vs your trap, and measure with judgment. Orient through the chapter, see the stack-map (hub + 6 spokes), and pick a recommended starting course based on your role and current pain

Map the 8 marketing jobs onto the 4 AI tool families (chat agents, research surfaces, embedded AI, automation), learn the 4-question routing rubric, and assemble the small multi-tool stack that actually fits a marketer's week — the foundation for the rest of the chapter

Workshop the 3 operational moves of brand-voice work with AI — extract 5-7 voice rules from 3-5 sample documents, capture them as a persistent system prompt (Custom GPT, Claude Project + Skill, or Gemini Gem) so every chat starts on-voice, and recognize the 5 drift signals that say it's time to reseed before shipping

Recognize the 4 moments where AI genuinely changes campaign planning (cited audience research, message-variant generation, pre-launch red-team, pre-ship verification) and the 3 dangers where its fluency makes things worse (positioning judgment, stakeholder politics, budget trade-offs) — operational moves with a citation discipline that keeps invented buyers out of your dossier

GEO is not better SEO. Different game, different metrics. Master the 4 operational moves — JSON-LD structured data AI engines prefer, per-engine brand-mention saturation (ChatGPT, Perplexity, AI Overviews each cite from different sources), citation-worthy content formatting (the opening 200 words rule + 5 more), and the hybrid SEO/GEO playbook — plus how to measure visibility with Otterly / Profound

Build 3 small marketing tools end-to-end with an AI coding agent (Claude Code as worked example; Cursor / Codex / Gemini CLI as escape hatches) — a competitor blog tracker from RSS feeds, a campaign data analyzer that turns HubSpot CSV exports into AI-narrated insights, and a content calendar generator that produces voice-matched multi-channel calendars. With the personal-stays-personal discipline that keeps these tools from becoming accidental products

The meta-skill that outlasts every tool churn — AI is useful for thinking, dangerous for deciding. 4 marketing cases (AI buyer personas, market sizing, competitive analysis, campaign ROI prediction) plus a 2-question decision rule (generate or adjudicate? if wrong, reversible or not?) that applies to any AI tool — including ones that don't exist yet

The chapter's analytical close — use AI to interpret analytics faster (compute outside AI, narrate with AI), design tighter A/B tests (hypothesis crispness + sample-size sanity + when to use multi-arm bandits), and recognize the 3 overclaim patterns AI is prone to (confounded variables, cherry-picked windows, vanity metrics) that turn fluent dashboards into bad decisions

Orientation across the AI for Sales stack — why AI is rewriting EVERY stage of the funnel in 2026, the seven-course map (five funnel-stage courses run linearly: prospecting → outreach → call prep → forecasting → proposals; plus two satellites — Coding-Lite for Sales and Microsoft Copilot for Sales — that cross-cut the funnel), the leverage map separating where AI is decisive (research speed, draft scaffolding, signal ranking, summarization) from where the human stays decisive (judgment, trust-building, accountability, verification), the concrete capabilities at stack end, and the pick-by-pain table that picks your starting course (default: sales-prospecting)

Bridge from knowledge-worker AI use to AI engineering for sales — reframe the "I'm not a developer" mental block (the new floor: reps who can ship a small custom tool are 10× more leveraged than reps who can't), install a coding agent once (Claude Code as the worked example with explicit escape hatch to Cursor / Codex CLI / Gemini CLI), learn the describe-codes-run-fix loop, ship three running mini-projects in your actual workflow (pipeline report extending sales-pipeline-forecasting, custom lead score extending sales-prospecting, outreach personalizer extending sales-outreach-scale Tier 2) with the EXACT reproducible prompts, recognize when NOT to use a coding agent (native CRM/Apollo feature exists, data too sensitive, silently-wrong-numbers in a board deck), and walk away with a coding-lite system — folder structure, three running tools, weekly cadence

Orientation across the AI Foundations skill path — vocabulary, mechanism, judgment, daily-practice habits, and prompt persistence, with judgment-depth chapters on reading and building AI benchmarks, plus five paths through based on where you're starting from

A decoder ring for every AI concept — what it means, which shelf it goes on, and the next question to ask

The machine under the magic — tokens, embeddings, attention, and prediction, explained without math

Five patterns, four surfaces, and a 7-day plan that turns ChatGPT or Claude into a daily co-worker by the end of week one

Spot the three failure modes, run a 30-second verify, and build the daily habit that keeps AI useful instead of dangerous

Stop retyping the same context every chat — configure ChatGPT Custom Instructions, Claude Projects, and Gemini Gems as your persistent prompt library, and recognize when a long chat has stopped serving you

Decode the percentages every vendor cites — what they measure, what they hide, and how to read a model report card without an engineering degree

Design and run a mini-benchmark on your own data to decide whether AI is really ready for a real task — no code required

A 12-minute orientation across the six-chapter BA skill path — from requirements elicitation to enterprise governance, the complete Copilot toolkit for advanced business analysis

Turn messy stakeholder inputs into structured BRDs, user stories, and acceptance criteria using M365 Copilot in Word, Loop, and Teams — with enterprise-scale context priming.

Use M365 Copilot in Visio, Whiteboard, and Excel to accelerate process documentation, build as-is/to-be comparison matrices, and surface cross-departmental inefficiencies.

Translate technical BA findings into audience-adapted communications using PowerPoint, Outlook, Loop, and Teams — managing competing interests and preserving precision across enterprise stakeholder levels.

Use Excel Copilot, Python in Excel, and Power BI Copilot to pivot, clean, analyze, and visualize enterprise data exports — building stakeholder-ready insights from messy CSVs without writing SQL or Python manually.

Trace how proposed changes ripple through systems, teams, and processes — building impact matrices, risk registers, and readiness assessments that synthesize technical and organizational dependencies for complex enterprise change programs.

Navigate sensitivity labels, DLP policies, data residency constraints, and audit requirements to remain productive with Copilot inside heavily regulated enterprise environments — from financial services Chinese walls to cross-border data boundaries.

Your orientation to creative AI foundations — landscape, inputs, collaboration, rights, and compound systems

Collaboration patterns that make human-AI creative work better than either alone

Build a creative asset system that compounds — every project makes the next one faster and better

Navigate ownership, Content Credentials, and the C2PA standard for AI-generated creative work

A clear-eyed map of AI creative tools in 2026 — what works, what is hype, and how to think about a stack that changes every quarter

How to prepare the inputs that determine whether AI creative tools produce magic or mediocrity — structured briefs, curated references, and living moodboards

A 12-minute orientation across the five-chapter Daily AI Toolkit — name the four AI tool families that the foundation chapter teaches, see how the three daily-surface chapters (docs/decks, email/meetings, research) apply that framework, recognize the ecosystem-routing chapter as the meta-decision overlay, and pick the right starting chapter based on which surface is biting hardest

Map the four AI tool families, build a default stack of three or four tools that fits your week, and learn the meta-skill of evaluating any new tool that lands

From a pile of messy inputs to a polished document and a stakeholder-ready deck — one repeatable 4-step pattern, executed across M365, Workspace, or hybrid stacks

Triage your inbox in minutes, capture meetings cleanly with the right notetaker, and run the post-meeting chain (transcript → summary → actions → per-person follow-up) so nothing falls out — with the privacy and consent layer that keeps it professional

Replace 12 browser tabs with cited answers in minutes — name the four research surfaces, brief deep-research runs that compress 4 hours of manual work into 25 minutes of yours, decide when a browser agent earns the click delegation, and apply the 3-source rule that keeps fabricated facts out of your work

Diagnose whether you are M365-native, Workspace-native, or hybrid — then route every AI task with rules that fit the ecosystem you actually live in, and know when to leave Chapter 2.1 for the deep dives (Copilot Mastery, future Workspace AI, AI Coding for Beginners)