Back to blog

How to Analyze a Lesson with AI

Why Analyze a Lesson?

Every teacher wants to improve their teaching, but objectively evaluating your own lesson is challenging. Self-reflection is subjective, and inviting a methodologist takes time and resources. AI analysis solves both problems: objective metrics in just a few minutes.

Step 1: Record Your Lesson

Use any device — smartphone, laptop, tablet, or surveillance camera. Etodist supports over 20 file formats: MP3, MP4, WAV, M4A, OGG, WebM, DAV, and more.

Tips for quality recording:

  • Place the microphone closer to the teacher
  • Minimize background noise
  • Record the entire lesson for complete analysis

Step 2: Upload the Recording

Log in to your dashboard at lk.etodist.app and click "Upload Lesson." Select your file, specify the subject, grade, and student age — this information helps the AI adapt its analysis to your specific context.

Step 3: Wait for Processing

The processing pipeline includes three stages:

  1. Audio extraction — if video is uploaded, FFmpeg extracts the audio track
  2. Transcription — Deepgram Nova-3 neural network recognizes speech with speaker diarization
  3. AI analysis — Claude analyzes the transcription across 25 pedagogical metrics

Processing takes 2–5 minutes for a one-hour lesson. Progress is displayed in real time.

Step 4: Review the Report

The completed report contains five tabs:

  • Overview — average score, key metrics, summary, and recommendations
  • Speech distribution — how much the teacher, students speak, and silence time
  • Teaching techniques — which methods were used: questioning, explanation, demonstration
  • Speech rate — pace of speech in 5-minute segments
  • Transcript — full lesson text with speaker separation

25 Metrics: What the AI Analyzes

The analysis is based on recognized pedagogical frameworks:

10 core metrics: Engagement, lesson structure, differentiation, feedback, pacing, clarity of explanation, use of questions, classroom management, lesson closure, use of visual aids.

15 pedagogical metrics: Bloom's Taxonomy, Vygotsky's Zone of Proximal Development, Growth Mindset (Dweck), formative assessment, metacognitive strategies, inclusivity, and more.

What to Do with the Results?

  1. Compare lessons with each other — track your progress over time
  2. Share reports with methodologists and colleagues via PDF export
  3. Focus on 2–3 metrics at a time for targeted growth
  4. Use AI recommendations as a starting point for professional development

Get Started Now

Your first analysis is free. Register and upload your first lesson today.