Allow users to upload recorded lecture audio or transcripts

[Nyumat]

Overview

Expand BeavsAI's capabilities by allowing users to upload recorded lecture audio or transcript files. This will enable the app to analyze and extract key information from lectures, providing students with an enriched learning experience through AI-powered insights and question generation.

Objective

Implement functionality that supports the upload and processing of audio files and transcript documents, making it easier for students to use lecture content as part of their study materials.

Details

Core Features

• File Upload: Extend the current upload feature to accept audio formats (e.g., MP3, WAV) and transcript documents (e.g., TXT, DOCX).
• Audio Transcription: Integrate a speech-to-text service to convert audio files into text for further processing.
• Content Parsing: Enhance existing parsing logic to handle transcribed audio and transcript documents for quiz generation and AI insights.
• UI Update: Update the upload page to allow users to choose between uploading PDFs, audio files, or transcripts.

Technical Requirements

• File Handling: Implement file parsing and conversion logic for audio and text files.
• Speech-to-Text Integration: Use a service like Whisper by OpenAI, AWS Transcribe, or similar for accurate transcription.
• Type Safety and Validations: Ensure type-safe code and Zod validations for file processing and error handling.
• Processing Time: Implement async processing for audio transcription to handle larger files without disrupting the user experience.

UI/UX Changes

• Upload Interface: Update the existing upload form to include options for audio and transcript file types.
• Progress Indicator: Provide feedback to users on the status of the transcription process.
• File Status: Notify users once the transcription and parsing are complete and available for analysis or quiz generation.

Data and Privacy Considerations

• Data Security: Ensure that uploaded audio and transcript files are processed securely and deleted after processing, if necessary.
• User Consent: Include a disclaimer informing users that their uploaded data will be analyzed by an AI and handled in compliance with data privacy regulations.

Acceptance Criteria

• Users can upload recorded lecture audio or transcript files successfully.
• Audio files are transcribed into text and parsed for insights.
• Transcription process is indicated in the UI, with status updates.
• Uploaded content is processed and available for quiz generation and AI insights.