Transform meeting transcripts into strategic assets—not soulless AI summaries, but actual insights you can learn from. Meeting Debrief + Effectiveness Review + optional Personal Subtext, all in under 3 minutes. Most meeting notes are transcription theater. “John mentioned the deadline.” “Team discussed options A and B.” Nobody learns from this. Meeting Intelligence transforms transcripts into three outputs: a Meeting Debrief (structured knowledge), an Effectiveness Review (neutral feedback), and optionally a Personal Subtext (private self-improvement notes).

Insight
We don’t have a meeting problem. We have a learning-from-meetings problem. AI can extract what matters—if you tell it what to look for.

Why “Debrief” Not “Notes”

“Meeting Notes” implies transcription—who said what, when. What we create is processed intelligence: topics extracted, insights identified, effectiveness reviewed.

The output is a debrief—military term for post-mission analysis. Not stenography.

The Problem

SymptomRoot Cause
Hours synthesizing notesManual processing doesn’t scale
Meetings drift from purposeNo feedback loop
Same dysfunctions repeatNobody reviews what went wrong
”That could have been an email”Wrong format for content

The Solution: 3 Outputs

1. Meeting Debrief

The shareable output for your team:

  • Decisions with owners
  • Action items with deadlines
  • 4-6 insights embedded in narrative
  • Cross-references to related meetings

2. Meeting Effectiveness Review

Insight
When a neutral tool says “this wasn’t a Daily, this drifted into problem-solving”—that has different impact than a person saying it. No blame, just data.

What AI evaluates:

  • Format fit: Daily vs. actual content
  • Time efficiency: 45-min topic in 15-min slot
  • Decision velocity: Made vs. deferred
  • Pattern detection: Same issues recurring

Deep Dive: Meeting Effectiveness Review →

3. Personal Subtext (Optional)

Private layer for self-improvement:

  • What could I have done better?
  • Patterns I’m repeating
  • Political dynamics I should be aware of

Deep Dive: Personal Subtext & Manöverkritik →

DSGVO Compliance

✅ What We DO❌ What We DON’T
Analyze meeting structureProfile participants
Extract decisions + actionsTrack “who said what”
Evaluate meeting formatEvaluate individual performance
Generate transferable insightsStore political subtext about people

Core principle: Evaluate the meeting, not the people.

Deep Dive: DSGVO-Compliant Meeting Analysis →

Why Orchestration (Not Single Prompt)?

Single PromptOrchestrated Phases
Too much context → LLM loses focusEach phase has manageable scope
No checkpoints → Errors cascadeValidation gates between phases
One failure → Entire output brokenErrors caught early

A 90-minute meeting transcript is too much for one prompt. Breaking it into phases (structure analysis → insight extraction → debrief generation → subtext generation) produces better results.

Getting Started

Minimum setup:

  1. Whisper transcription (local or API)
  2. Claude with structured prompt
  3. 3 minutes per meeting

Full setup (what I use):

  • Whisper.cpp local transcription
  • Claude Code with orchestrated commands
  • Obsidian for note storage + linking

Deep Dives

  1. Example: Meeting Debrief + Subtext — Full example of debrief + private subtext
  2. Meeting Effectiveness Review — The neutral feedback loop for teams
  3. Personal Subtext — Private Manöverkritik for self-improvement
  4. DSGVO Compliance — What’s allowed, what’s not, how to stay safe

Sources

  • Personal experience: 100+ meetings analyzed with this system
  • Whisper: OpenAI’s speech recognition model
  • Spacing effect research: Insights distributed in narrative vs. summary at end
  • Doc patterns: Documentation Patterns for insight callout format

Deep Dives

01

Example: Meeting Debrief + Subtext

Full example of a Meeting Debrief (shareable) and Personal Subtext (private) from a real onboarding session.

02

Meeting Effectiveness Review: Neutral Feedback for Teams

When a neutral tool says 'this wasn't a Daily'—that lands differently than when a person says it. AI becomes the honest mirror your team needs.

03

Personal Subtext: Private Manöverkritik

The official debrief is for the team. The subtext is for you—honest feedback on what you could do better, patterns you're repeating, and things you need to know.

04

DSGVO-Compliant Meeting Analysis

You can analyze meetings with AI—but not people. Focus on structure, decisions, and format. Never on individuals.