AI / Productivity

AI NoteTracker

FlutterFlow Application

Preview coming soon

Overview

AI NoteTracker turns spoken audio into organized, editable text notes. Users record on the spot or upload existing audio, and the app runs it through AI-powered transcription and files the result into cloud folders alongside their tasks.

The Challenge

Audio-to-text conversion introduces asynchronous processing time that the UI needs to represent clearly, and the resulting notes still needed to live inside an organized cloud folder structure rather than as a flat list.

My Role

I built the application in FlutterFlow with Firebase as the backend, implementing the audio capture and upload flow, the AI transcription integration, and the cloud folder and task management system.

Technical Approach

Built an audio capture flow supporting both live recording and file upload, feeding either source into the same transcription pipeline.
Integrated AI-powered audio-to-text conversion and surfaced clear in-progress and completion states while transcription runs.
Implemented cloud-based folders in Firebase Storage so generated notes, uploaded files, and music sit in a structure the user organizes themselves.
Added task and to-do management alongside notes so the app covers both capture and follow-up action.
Built music upload and playback as a secondary content type sharing the same cloud folder system.

Technical Challenges & Solutions

Transcription isn't instant, and users needed to trust that a recording wasn't lost while processing.

Solution — Persisted the raw audio immediately on capture and updated the note's status once transcription completed, so the source was never dependent on a successful AI response.

Mixing notes, tasks, music, and arbitrary files in one storage system risked becoming disorganized as content volume grew.

Solution — Structured Firebase Storage around user-created folders with typed file references, keeping each content type queryable independently.

Technical Architecture

Built in FlutterFlow with Firebase as the backend — Firestore for notes and task metadata, Firebase Storage for audio, music, and file folders. Audio-to-text runs as an asynchronous AI pipeline: raw audio persists immediately on capture, and the note record updates in place once transcription completes, so the source is never dependent on a successful AI response.

Outcome

Delivered a working AI-powered note application that reliably converts recorded and uploaded audio into text, organized alongside tasks and files in cloud folders.