Tech & AIBeginnerPreview
Notion AI for Productivity
A hands-on course that takes you from typing space-bar prompts to running an AI-assisted workspace: summarize long documents, auto-fill database fields with AI properties, draft polished content blocks, and stand up a wiki and a project hub that stay current with less manual effort. You leave with reusable AI-property setups, a prompt library, and at least one live hub for your team.
For operators, founders, project managers, marketers, researchers, and team leads who already use Notion and want practical, repeatable productivity gains from its AI features.
Course content
Workbook & downloads
Put the course into practice — a printable workbook plus editable templates you can fill in and reuse.
Preview the workbook
This workbook turns the course into a working AI-assisted Notion setup. Work through one section per module: get fluent invoking AI, build a reusable summary and extraction toolkit, stand up a self-maintaining database, then assemble an AI-grounded wiki and project hub with sane guardrails. Fill the worksheets directly in Notion, run each build step in your real workspace, and by the end you will have a prompt library, at least one AI-powered database, and a live hub your team can ask questions of.
Notion AI Foundations and Setup
Confirm access, learn the four AI surfaces, run your first prompts well, and decide where AI is worth using.
Checklist: Access and Surfaces Checklist
- Confirm your plan (Plus, Business, or Enterprise) includes Notion AI in Settings
- On an empty line, press the space bar and confirm the Ask AI menu appears
- Open AI Q&A with Cmd or Ctrl plus J and confirm the ask box opens
- Highlight text and confirm the Ask AI button appears for transforming a selection
- Type slash then ai and note the available commands (Summarize, Translate, Improve writing)
- Pick one real document and one real database to practice on for the whole course
Worksheet: First Prompt Builder
Draft three in-page prompts for tasks you actually do, filling each row. Keep prompts to one or two sentences and always name the output shape and any guardrail. You will run these in Notion and keep the best ones in your library.
- Prompt 1 — task verb (summarize, extract, rewrite, draft)
- Prompt 1 — output shape (bullets, two sentences, table)
- Prompt 1 — guardrail (do not invent figures, keep all dates)
- Prompt 2 — task verb
- Prompt 2 — output shape
- Prompt 2 — guardrail
- Prompt 3 — task verb
- Prompt 3 — output shape and guardrail
Exercise: Use-or-Skip Decision
For one task you are tempted to hand to AI, answer each prompt in a single line. If you answer no to the second prompt or yes to the last, keep a human review step or do it manually.
- Will this task repeat, and could someone else need the result? Explain in one line.
- Can I verify the AI output against a source in under a minute? How?
- Does the source material already live in my workspace, or must I paste it in?
- Would a mistake here cost money or embarrass me externally? If so, what is my review step?
Summarizing and Extracting from Documents
Build a reusable toolkit for summaries, action-item extraction, and reshaping existing content.
Worksheet: Summary Shape Selector
For a real long document, plan two summaries that serve two different readers. Fill both columns, then generate each in Notion and pin them to the top of the page.
- Document name and rough length
- Reader A and what they need (e.g. executive, two-sentence so-what)
- Summary A prompt (shape, length cap, guardrail)
- Reader B and what they need (e.g. team, five-bullet skim)
- Summary B prompt (shape, length cap, guardrail)
- Two or three load-bearing facts to spot-check against the source
Exercise: Action-Item Extraction Drill
Paste a real transcript or set of meeting notes onto a page, then run the three-section extraction prompt. Answer the prompts below about the result to confirm it is actually usable.
- Did every action item come out in owner colon task form, with Unassigned where no owner was named?
- Which task surfaced as Unassigned that you would have otherwise missed?
- Are the decisions stated as completed decisions, and are any invented (not in the source)?
- What one tweak to the prompt would make next week's extraction cleaner?
Worksheet: Repurpose Planner
Pick one existing document and plan how to reshape it for a new audience without rewriting from scratch. Fill the fields, then run the transform in Notion and review the output.
- Source document (what it is, who it was for)
- New format and new audience (e.g. spec to customer FAQ)
- Reshape prompt (structure, length, what to keep or omit)
- Fixed terms to preserve (product names, acronyms, English-only terms)
- Review step before this goes anywhere external
Checklist: Document Toolkit Checklist
- Saved at least two summary prompts (executive and skim) to a prompt library page
- Saved the three-section action-item extraction prompt to the library
- Saved one reshape or translate prompt to the library
- Verified key figures in one real summary against the source text
- Converted one extraction into a working to-do checklist with an owner assigned
AI-Powered Databases
Configure AI properties that classify, summarize, and extract so a database maintains itself.
Worksheet: AI Property Plan
Plan the AI columns for your chosen database before you build them. Fill one block per property. Keep classification labels to a short closed list and always include a fallback label.
- Database name and what each row represents
- Property 1 type (Summary, Key Info, Translation, or Custom Autofill)
- Property 1 prompt and which field it reads
- Property 2 type
- Property 2 prompt, allowed labels (closed list), and fallback label
- Update mode per property (on-demand, automatic, or periodic bulk fill)
- View to build (board or table grouped by the classification column)
Exercise: Classification Prompt Tightening
Build a custom AI classification column, backfill ten rows, then audit the labels. Use the prompts to diagnose and tighten the instruction until labels are consistent.
- Of ten backfilled rows, how many got a label outside your intended list or a wrong call?
- Did the model invent any label instead of using your fallback? What was it?
- Which field should the prompt read to be most accurate, and did you specify it?
- Rewrite the closed list or fallback wording to fix the errors you found.
Checklist: Self-Maintaining Database Checklist
- Added at least one AI Summary or Key Info column with a precise output target
- Added a custom classification column with a closed label list and a fallback
- Told each AI property which field(s) to read
- Backfilled existing rows once, then spot-checked ten cells for accuracy
- Built a gallery or board view that shows gist, key fact, and category together
- Set a deliberate update mode per column instead of auto-updating everything
AI-Assisted Wikis and Project Hubs
Build a wiki and a project hub Notion AI Q&A can answer from, with accuracy, access, and cost guardrails.
Worksheet: Wiki Source-of-Truth Map
Plan a wiki that AI Q&A can answer from. List the authoritative pages, their owners, and any duplicates to merge. One source of truth per topic is the rule.
- Wiki name and audience
- Topic 1 — page title, owner, verified yes/no
- Topic 2 — page title, owner, verified yes/no
- Topic 3 — page title, owner, verified yes/no
- Duplicate or stale pages to merge or archive
- Three real questions you will test against Q&A
Exercise: Project Hub Q&A Test
Build a project hub (overview, linked tasks database, notes database with AI Summary) and connect one external source if you can. Then ask AI Q&A the questions below and judge the answers.
- Ask what the open or blocked tasks are. Did Q&A answer from the tasks database and cite it?
- Ask what was decided about a specific topic. Was the answer grounded in your notes with a citation?
- Click one citation. Did the source page actually support the answer?
- What is missing or stale in the hub that caused a weak answer, and how will you fix it?
Worksheet: Guardrails Worksheet
Set your team's guardrails for accuracy, access, and cost. Fill each field; these become norms you share on the hub.
- Verification rule (when must someone click the citation before acting?)
- Which connectors are enabled and who approved them
- Plan data terms note (is your content excluded from external model training?)
- AI-property cost norm (when is automatic update allowed on shared databases?)
- Owner of the monthly wiki re-verification and database audit
Checklist: Go-Live and Maintenance Checklist
- Wiki has three or more authoritative pages, each with an owner and verified status
- Duplicate and stale pages archived or merged so Q&A does not conflict itself
- Project hub links an overview, a tasks database, and a notes database with AI summaries
- AI Q&A answers three real questions with correct citations
- Verification, connector, and cost guardrails written on a shared norms page
- Monthly re-verify and audit reminder scheduled, prompt library shared with the team
Your Action Plan
- Confirm Notion AI access and practice all four surfaces: space-bar prompt, highlight-then-ask, AI Q&A, and AI properties.
- Create a My AI Prompts library page and seed it with the starter prompts plus the ones you wrote.
- Generate two audience-shaped summaries (executive and skim) on a real long document and verify the key figures.
- Run the three-section action-item extraction on a transcript and convert the results into an assigned to-do list.
- Reshape one existing document into a new format for a new audience instead of writing it from scratch.
- Add AI Summary and Key Info columns plus a custom classification column to one database, then backfill and audit.
- Set a deliberate update mode per AI property and build a grouped view that reads like a dashboard.
- Build a wiki of three to five owned, verified, authoritative pages and archive duplicates.
- Assemble a project hub with linked tasks and notes databases, connect one external source, and test AI Q&A with citations.
- Write accuracy, access, and cost guardrails on a shared page and schedule a monthly re-verify and audit.
Pairs well with
Courses members commonly take alongside this one.
Flagship CoursePreview
Freelance Business Foundations: Position, Price, Sell, and Deliver High-Value Services
Freelancing · Beginner · 16h
Self-pacedPreview
Client GrowthPreview
Freelance Client Acquisition: Outreach, Leads, Referrals, and Deal Flow
Freelancing · Beginner · 15h 30m
Self-pacedPreview
Sales SystemPreview
Freelance Sales & Proposals: Discovery Calls, Scoping, Objections, and Closing
Freelancing · Intermediate · 16h
Self-pacedPreview