What Are Smart Notes?
Smart Notes is a feature that uses AI to analyze your completed transcription and automatically create structured markers at meaningful moments throughout your recording. Unlike simple time-based segmentation that divides audio into equal five-minute chunks, Smart Notes identify actual content boundaries — where a new topic begins, where a key definition is given, where an argument reaches its conclusion, where a decision is made.
The result is a set of markers that reads like an intelligent table of contents for your recording. Each AI-generated marker has a descriptive title (like "Definition of thermodynamic equilibrium" or "Decision: launch date moved to Q2"), a precise timestamp linking to the exact moment in the audio, and often a summary or key quote extracted from the transcript.
You can tap any Smart Notes marker to jump directly to that moment in the recording, read the surrounding transcript text, or use it as a navigation point during study or review sessions. Smart Notes transform a raw audio recording into structured, navigable knowledge.
How They Differ from Manual Markers
Manual markers and Smart Notes serve complementary purposes, and understanding the difference helps you get the most from both.
Manual markers capture what you noticed during the live recording. They reflect your attention, your priorities, and your real-time judgment about what felt important while you were listening. Manual markers are inherently subjective — they represent your perspective on the content.
Smart Notes capture what the content contains based on structural and semantic analysis. The AI reads the entire transcript, identifies topic transitions, key statements, definitions, arguments, decisions, and conclusions, then creates markers at each of these structural points. Smart Notes are comprehensive — they cover the entire recording, including parts where you were distracted, taking notes on paper, or simply did not notice something important.
These two approaches fill each other's gaps perfectly. Your manual markers flag the moments that resonated with you personally — the insight that changed your thinking, the quote that answered your question. Smart Notes fill in the structural framework — the topic outline, the supporting arguments, the context around your manual markers that helps you understand why those moments mattered.
How to Generate Smart Notes
Smart Notes require a completed transcription as their input. The AI analyzes text, not audio, so the transcription step must finish before Smart Notes generation can begin. Here is the complete process from recording to structured notes:
- Record your audio session inside a project, or import an existing audio file from another source
- Transcribe the recording — either by tapping Transcribe manually or through auto-transcription. Wait for the transcript to reach Completed status.
- Once transcription is complete, open the project and tap Generate Smart Notes. The button appears only after a completed transcript is available.
- Wait 1-3 minutes depending on the recording length. During this time, the AI reads the full transcript, identifies key structural elements, and creates markers with descriptive titles at precise timestamps.
- When finished, the AI-generated markers appear in your marker list alongside any manual markers you created during recording. They are visually distinguished but function identically.
Smart Notes are fully editable after generation. You can rename them to better match your terminology, delete ones that are not relevant to your needs, change their marker type (Flag, Favorite, Fire), adjust their timestamps, or add your own descriptions. Once created, they behave exactly like manual markers in every way.
Quality Factors
The quality and usefulness of Smart Notes depends on several factors that you can influence:
- Transcription accuracy — Smart Notes analyze the transcript text, so transcription quality directly affects the result. Better audio quality leads to better transcription, which leads to more accurate and relevant Smart Notes. See How Transcription Works for tips on improving accuracy.
- Content structure — lectures and presentations with clear topic transitions, explicit definitions, and organized arguments produce the most useful Smart Notes. Free-form conversations and brainstorming sessions still work, but the markers reflect the conversational flow rather than a strict topic outline.
- Recording length — longer recordings give the AI more context to work with, generally producing better structural analysis. A 5-minute voice memo has less structure for the AI to identify than a 60-minute lecture.
- Project type setting — choosing the correct project type (Lecture, Meeting, Other) is important because it tells the AI what kind of content patterns to look for. A Lecture type looks for definitions and arguments. A Meeting type looks for tasks and decisions. See Project Types for details.
Manual + AI Combined
The most effective workflow combines both manual markers and AI Smart Notes into a single, comprehensive review system:
- During recording, add manual markers whenever something stands out — do not worry about being thorough or systematic, just capture what feels important in the moment
- After the recording is transcribed, generate Smart Notes to get the AI's comprehensive structural analysis of the entire content
- Review the combined marker list — your personal highlights sitting alongside the AI's systematic topic markers
- Edit, rename, and reclassify the markers into your final organized notes. Upgrade the most important ones to Favorite or Fire. Add descriptions with context and study notes.
This combined approach ensures that nothing important is missed. Your intuition captures the moments that matter to you personally, and the AI captures the structural framework and contextual details that connect everything together. The result is significantly more complete than either approach alone.