Notion AI Pocket Book

Posted on by uplatzblog

Notion AI Pocket Book — Uplatz

50 Expanded Cards • Prompts • Workflows • DB automations • Integrations • 20 Interview Q&A

1) What is Notion AI?

Notion AI is a set of AI-assisted features built into Notion that helps you write, summarize, translate, brainstorm, transform text, and automate content creation directly inside pages and databases.

Use: Space → Ask AI  |  Slash: /Ask AI
Context-aware: Summarize selected text or entire page.

2) AI Quick Actions

Trigger inline: Improve writing, Fix spelling & grammar, Make longer/shorter, Change tone. Great for polishing docs and meeting notes.

Select text → "Ask AI" → "Change tone: confident" → Review diff

3) AI Blocks & Slash Commands

Use slash to insert AI blocks that generate content on demand. Keep prompts near the content for reproducibility.

/ai: "Create a 5-bullet summary of this page for executives."

4) Context Windows & Scope

Notion AI uses the current page, selected blocks, and linked databases as context. More structure → better outputs.

  • Use headers and callouts to delineate sections.
  • Link database rows with @-mentions to bring structured context.

5) Prompt Patterns

Be explicit about audience, format, constraints, and examples (few-shot). Iterate prompts in the same block.

Prompt: "You are a PM. Summarize risks in 5 bullets. Include owner & due date."

6) Page Summarization

Quickly generate executive summaries or “tl;dr”. Ideal for weekly updates and decision logs.

Action: Ask AI → "Summarize page with action items & deadlines"

7) Brainstorm & Outline

Kickstart new docs with outlines, angles, and counterarguments. Convert outline → doc with headings.

"Brainstorm 8 onboarding email ideas for B2B SaaS → cluster by theme"

8) Rewrite by Tone

Transform a block into academic, friendly, concise, or persuasive tone. Use to tailor for stakeholders.

Highlight paragraph → "Ask AI → Make concise + professional"

9) Translate & Localize

Translate selected blocks or pages to reach global audiences. Add style guidance: “Keep product names English.”

Ask AI → "Translate to Spanish. Preserve code fences and UI labels."

10) Tables & Databases + AI

Use AI to fill, clean, or summarize database fields. Great for auto-tagging, extracting entities, and generating descriptions.

In DB row → Ask AI: "Generate a 120-char SEO description for this feature."

11) Meeting Notes → Actions

Turn messy notes into action lists and decisions. Keep owners & due dates consistent.

Select notes → Ask AI: "Extract action items: [Task | Owner | Due | Status]"

12) Email & Outreach Drafts

Generate personalized outreach from CRM properties stored in a Notion database.

Prompt: "Draft a follow-up email. Use {{First Name}}, reference {{Last Call Summary}}."

13) Product Specs from Templates

Maintain a spec template and ask AI to fill summary, scope, success metrics from page content.

"Fill the Spec Template sections using headings: Summary/Goals/Out of scope"

14) Research Digest

Paste research notes, then ask AI for key insights, quotes, and source table.

"Create a table: Source | Quote | Insight | Confidence (Low/Med/High)"

15) Content Calendar Ideas

Seed a content DB with AI-generated titles, briefs, and SEO keywords.

"Generate 12 blog titles for {{ICP}}. Include target keyword & funnel stage."

16) Policy & Docs Simplification

Convert dense policy text into simple FAQs and checklists for non-technical teams.

"Turn this security policy into a 10-question FAQ. Include yes/no checkboxes."

17) Learning Notes → Flashcards

Create Q/A blocks from lecture notes or book highlights.

"Extract 20 flashcards: Question | Answer | Difficulty (1-3)"

18) Code Review Summaries

Paste PR description & diffs — AI generates risk callouts, test ideas, and release notes.

"Summarize this PR: risks, backwards-compat, test checklist, release notes."

19) Slide Narrative Builder

Draft slide outlines and speaker notes from a one-paragraph idea.

"Create a 10-slide outline with titles + 3 bullets + speaker notes."

20) Job Description & Scorecards

Standardize JDs and interview scorecards tied to competencies in a Notion DB.

"Draft JD for {{Role}}. Add responsibilities, must-haves, 5 scorecard criteria."

21) Classification & Tagging

Use AI to categorize content into tags (topic, stage, persona). Store as Multi-Select.

"From text, infer tags: {Persona, Stage, Topic}. Return comma-separated."

22) Entity Extraction

Pull companies, people, dates into relation properties for cross-linking.

"Extract: Organizations, People, Dates → list JSON arrays only."

23) Summaries per Row

Generate 1–2 sentence summaries for each database item, ideal for galleries & cards.

"Summarize this row in 30 words max. Start with a strong verb."

24) Consistent Naming Conventions

Normalize titles: add prefixes, ticket references, or version numbers using AI + formulas.

Formula: "TP-" + slice(prop("ID"), -4) + " · " + title

25) Generate Acceptance Criteria

From user stories, produce testable acceptance criteria into a checklist property.

"From user story, output 6 Given-When-Then AC bullets."

26) Risk Register Automation

Convert notes into Risk entries with Severity, Likelihood, Owner.

"Create risk: Title, Impact, Likelihood(1-5), Mitigation, Owner, Due"

27) CRM Enrichment

Turn raw notes into lead summaries, next best actions, and email drafts stored per record.

"Summarize call: pain points, budget, authority, next step (date + CTA)"

28) Roadmap Rollups

Roll up child task summaries into parent epic pages. Ask AI to condense.

"Summarize linked tasks (Status=In Progress) into 4 bullets."

29) Knowledge Base Normalization

Migrate disparate docs into a unified schema: Title, Audience, Version, Owner, Summary.

"Rewrite page to KB style: overview, steps, troubleshooting, owner."

30) SEO Snippets in DB

Generate meta titles, descriptions, and FAQs from article drafts.

"Produce: Meta Title(60c), Meta Desc(155c), 3 FAQs with answers."

31) Notion API Basics

Use the Notion API to create/update pages and query databases. Combine with AI outputs stored in properties.

# Pseudocode (JS)
notion.pages.create({ parent:{database_id}, properties:{ Name:{title:[{text:{content}}]}})

32) Zapier/Make Scenarios

Trigger: New DB item → AI summarize → Update “Summary” property. Or receive form submissions → generate draft page.

Zap: Typeform → Notion (create) → Notion AI (summary) → Slack

33) Slack + Notion AI

Pipe Slack threads into Notion; ask AI to condense into decisions & tasks.

"Summarize Slack thread: decisions, blockers, owners. 5 bullets."

34) GitHub → Release Notes

Sync PR titles to Notion; AI groups by features/bugfixes and drafts release notes.

"Group PRs by label. Write human-readable release notes."

35) Calendar Planning

From events DB, AI writes week plan and focus blocks. Send to Slack daily.

"Create a weekly plan from Events between Mon–Fri. Include timeboxing."

36) Support Tickets Triage

Import tickets; AI classifies severity, suggests responses, links KB articles.

"Classify ticket severity 1-4; propose reply; link to matching KB."

37) Finance & FP&A Notes

From CSV uploads, AI produces variance commentary and highlights anomalies.

"Explain MoM variance > 10%. Summarize 3 key drivers. Next actions."

38) HR Onboarding Flows

AI builds personalized onboarding checklists from role templates.

"Generate onboarding plan for {{Role}} week 1–4. Include buddies & goals."

39) Education: Course Outlines

Create course syllabi, readings, and assessments from a topic brief.

"Design 6-week syllabus for {{Topic}}. Weekly outcomes + assignments."

40) Personal Productivity OS

Use AI to maintain Tasks, Habits, and Journal databases. Auto-summarize days & weeks.

"Summarize today: wins, blockers, priorities tomorrow (3 items)."

41) Prompt Engineering Tips

  • Specify role, audience, format, length.
  • Provide 1–2 examples (few-shot).
  • Constrain outputs (tables, JSON).
"You are a tech writer. Output: 5 bullets, each < 15 words."

42) Reusable Prompt Library

Maintain a “Prompt Bank” DB with fields: Use Case, Prompt, Inputs, Output Format, Quality Notes.

Properties: Title | Use Case | Prompt | Vars | Example Output

43) Guardrails & Review

Institute human-in-the-loop checks: AI drafts → human review → publish. Track edits in a “Changes” property.

Workflow: Draft (AI) → Review (Human) → Approved → Published

44) Style Guides & Voice

Create a Voice page; instruct AI to follow it. Include do/don’t examples.

"Rewrite in 'Uplatz' voice: clear, friendly, actionable. Avoid jargon."

45) Privacy & Data Handling

Avoid sensitive data. Redact personally identifiable information. Store secrets in a separate, access-controlled DB.

Policy: No secrets in prompts. Use placeholders like {{API_KEY}}.

46) Cost & Performance

Batch operations, keep prompts short, and reuse summaries. Archive old pages to reduce context noise.

Tip: Summarize large pages once, then refine summaries.

47) Troubleshooting Poor Outputs

  • Clarify instructions & constraints.
  • Provide examples and counterexamples.
  • Reduce irrelevant context.
Debug Prompt: "List assumptions you used, then answer."

48) Certification-Style Practice

Simulate tasks: “Given a messy page, produce a structured brief + tasks.” Evaluate with rubric.

Rubric: Structure(30) Clarity(30) Actionability(20) Tone(20)

49) Interview Q&A — 10 Essentials

Q1: How does Notion AI use context? A: Current selection/page/linked DB.

Q2: Top prompt elements? A: Role, audience, format, constraints, examples.

Q3: DB use cases? A: Summaries, tagging, extraction, checklists.

Q4: Guardrails? A: HITL reviews, style guides, redaction.

Q5: Integrations? A: API, Zapier/Make, Slack, GitHub.

Q6: Improve quality? A: Few-shot, explicit outputs, iterate.

Q7: Risks? A: Hallucinations, privacy, bias. Mitigate with review.

Q8: SEO workflow? A: Titles, metas, FAQs from draft DB.

Q9: Research digest? A: Quotes → insights table.

Q10: Summarization limits? A: Break long docs; use section headers.

50) Interview Q&A — 10 Advanced

Q11: Prompt library design? A: DB with versioning, examples, tags.

Q12: Enforcing voice? A: Style page + few-shot examples.

Q13: API workflow? A: Create page → AI fields → status updates.

Q14: Large-team governance? A: Permissions, templates, review gates.

Q15: KPI ideas? A: Time saved, drafts accepted, edit ratio.

Q16: Handling hallucinations? A: Cite sources, require evidence blocks.

Q17: Multi-lingual docs? A: Language DB, translation prompts.

Q18: Data extraction accuracy? A: Output JSON schemas + validation.

Q19: Migrating legacy docs? A: Normalize templates + AI rewrite.

Q20: Scale content ops? A: Batch generate → human review → publish pipeline.

Example JSON schema:
{ "title": "", "summary": "", "tags": [], "owner": "", "due": "YYYY-MM-DD" }