Product Requirements Document (PRD): Fast Learner

Version: 1.0
Date: 2025-10-23
Author: (Your Name)

1. Overview

Fast Learner is a web application designed to help users learn more efficiently from YouTube videos. It transforms passive video consumption into an active learning experience by providing AI-powered summaries, interactive transcripts, translation, and focused playback controls. This tool is for students, professionals, and anyone who wants to extract key information from videos quickly without watching them in their entirety.

2. Target Audience

• Students: Researching topics, reviewing lectures, or studying for exams.
• Professionals: Staying updated on industry trends, learning new skills, or reviewing conference talks.
• Lifelong Learners: Exploring new hobbies or complex topics efficiently.

3. Core Features & User Stories

This section outlines the primary features based on the initial design, framed as user stories.

ID	Feature	User Story
F1	Video Ingestion	As a user, I want to paste a YouTube URL and click "Learn" so I can load the video and its learning tools.
F2	Video Playback	As a user, I want to watch the video with standard playback controls, including adjustable speed, so I can consume content at my preferred pace.
F3	Interactive Transcript	As a user, I want to see a full, time-synced transcript of the video so I can read along, search for keywords, and click any line to jump to that point in the video.
F4	Transcript Translation	As a user, I want to translate the transcript into my preferred language so I can understand content that isn't in my native tongue.
F5	AI Summary	As a user, I want to see an AI-generated summary of the video's key takeaways so I can get the main points at a glance.
F6	AI Q&A	As a user, I want to ask specific questions about the video's content so I can clarify my understanding and get deeper insights.
F7	Focus Mode	As a user, I want to enable a "Focus Mode" that hides distracting UI elements so I can concentrate on the video and text.
F8	AI Assist Toggle	As a user, I want to be able to toggle all "AI Assist" features (Summary, Q&A) on or off to control my experience and manage API usage.

4. Feature Specifications

4.1. Video Ingestion (F1)

• UI: A single text input field and a "LEARN" button (as sketched).
• Functionality:
  • Must accept full YouTube URLs (e.g., https://www.youtube.com/watch?v=...) and shortened links (e.g., https://youtu.be/...).
  • On "LEARN" click:
    1. Validate the URL. Show a simple error message (e.g., "Invalid YouTube URL") if it's not a valid YouTube link.
    2. Clear the previous video and its data.
    3. Load the video into the player (F2).
    4. Fetch the video's transcript (F3).
    5. If "AI Assist" is on, trigger the AI Summary (F5).
  • A loading indicator should be displayed while fetching data.

4.2. Video Player & Speed Control (F2)

• UI: An embedded YouTube IFrame player (as sketched in box '2').
• Functionality:
  • Use the YouTube IFrame Player API.
  • A "SPEED" control (as sketched in '3') should be present outside the player for easy access.
  • Suggested Speed Options: 0.5x, 0.75x, 1x (default), 1.25x, 1.5x, 1.75x, 2x.

4.3. Interactive Transcript Panel (F3)

• UI: A scrollable panel (as sketched in '4') located next to or below the video player.
• Functionality:
  • Automatically fetch the English transcript (or the default available language) for the video.
  • If no transcript is available, display a clear message: "Sorry, no transcript is available for this video."
  • Timestamp-Sync: As the video plays, the transcript panel should automatically scroll and highlight the currently spoken line.
  • Click-to-Seek: Each line/paragraph in the transcript must be clickable. Clicking a line seeks the video player to that specific timestamp.

4.4. Transcript Translation (F4)

• UI: A "Translate" button and a language-selector dropdown, placed near the top of the transcript panel.
• Functionality:
  • The dropdown should list major languages (e.g., Spanish, French, German, Mandarin, Japanese, Hindi).
  • When a user selects a language and clicks "Translate":
    1. Call the Gemini API.
    2. System Prompt: "You are a helpful translator. Translate the following text into [Selected Language]."
    3. User Prompt: [Full Transcript Text]
    4. Replace the transcript panel's content with the translated text.
  • A "Show Original" button should appear to revert to the original transcript.
  • Note: Translated text will lose its timestamp-sync functionality. This is an acceptable trade-off for V1.

4.5. AI Assist Panel (F5, F6, F8)

This panel (sketched as '6') combines the Summary and Q&A features, all governed by the "AI Assist" toggle (as sketched as '5').

• "AI Assist" Toggle (F8):
  • A master checkbox (as sketched).
  • If OFF, the AI Assist Panel (F5, F6) is hidden or disabled.
  • If ON, the panel is visible and functional. This should be ON by default.
• AI Summary (F5):
  • UI: The default view of the panel (sketched as "Summary A").
  • Functionality: When a video is first loaded (and AI Assist is ON), make a call to the Gemini API.
  • System Prompt: "You are an expert academic assistant. Based on the following video transcript, provide a concise summary of the video's key takeaways in 5-7 bullet points."
  • User Prompt: [Full Transcript Text]
  • The resulting summary is displayed in the panel.
• AI Q&A (F6):
  • UI: The '+' button (sketched in '6') should reveal a text input field at the bottom of the panel with a "Ask" button.
  • Functionality: The user can ask questions about the video.
  • System Prompt: "You are a expert assistant. The user is watching a video with the following transcript. Answer their question based only on the provided transcript context."
  • User Prompt: Transcript: [Full Transcript Text]\n\GQuestion: [User's Question]
  • The answer is appended to the panel, creating a chat-like history below the initial summary.

4.6. Focus Mode (F7)

• UI: A toggle button/checkbox labeled "Focus mode" (as sketched in '5').
• Functionality:
  • When ON, the UI is simplified to minimize distractions.
  • This mode should hide:
    • The app header/title.
    • The initial URL input bar.
    • The AI Assist Panel.
  • The Video Player and the Interactive Transcript panel should expand to fill the screen, side-by-side or stacked, depending on viewport size.
  • The "Focus mode" toggle itself should remain visible or be accessible via an icon to exit the mode.

5. Non-Functional Requirements

• Technology Stack (Suggestion):
  • Front-End: A single, self-contained index.html file.
  • Styling: Use Tailwind CSS via the CDN link for rapid, clean, and responsive UI development.
  • JavaScript: Vanilla JavaScript (ES6+). Use fetch() for all API calls.
• APIs:
  • YouTube: YouTube IFrame Player API (for embedding and controlling the player).
  • Transcript: Requires a method to fetch YouTube transcripts (e.g., via a simple serverless function or a third-party API, as YouTube doesn't provide a direct, unauthenticated transcript API).
  • AI: Google Gemini API (for Summary, Q&A, and Translation).
• Design:
  • Clean, modern, and minimalist.
  • Fully responsive: The layout must adapt seamlessly from mobile (stacked player/transcript) to desktop (side-by-side) views.
• Error Handling:
  • Display user-friendly messages for:
    • Invalid YouTube URL.
    • Video with no available transcript.
    • AI API call failures (e.g., "AI assistant is currently unavailable").

6. Future Ideas (V2 - Out of Scope for V1)

• User Accounts: To save video history, notes, and AI chat logs.
• Note-Taking: A dedicated panel for users to write their own timestamped notes.
• Export: Ability to export AI summaries, Q&A chats, and user notes as Markdown or PDF.
• Support for other platforms: Add support for Vimeo, Dailymotion, etc.