System Requirements Document (SRD) for Divine-Automation

1. Introduction

The Divine-Automation project is a cutting-edge automated video editing platform designed to streamline the video editing process for users in India and beyond. By leveraging advanced AI models and a robust backend architecture, this system automates tasks such as silence removal, intelligent scene cuts, and subtitle generation. This document outlines the system's requirements, ensuring clarity and alignment with the project's goals.

2. System Overview

Divine-Automation is an end-to-end automated video editing solution aimed at both casual users creating social media content and professionals requiring advanced editing capabilities. The system processes uploaded videos by applying a series of automated transformations, including:

• Silence removal to enhance audio clarity.
• Scene detection and intelligent cuts for seamless transitions.
• AI-powered subtitle generation for accessibility and engagement.

The platform consists of a frontend built with Next.js for user interaction, a backend powered by Node.js and Python for processing, and a Redis-based queue system to manage tasks efficiently. The system also integrates AI models like OpenAI's Whisper for transcription and subtitle generation.

3. Functional Requirements

User Stories:

• As a User, I should be able to upload a video file for processing.
• As a User, I should be notified when the video processing starts and completes.
• As a User, I should be able to download the processed video with subtitles.
• As a User, I should be able to preview the processed video before downloading.
• As a User, I should be able to select specific processing options (e.g., silence removal, scene cuts, subtitles).
• As an Admin, I should be able to monitor the processing queue and system performance.
• As a Guest, I should be able to view a demo of the platform's capabilities without uploading a video.

4. User Personas

1. Casual User

• Description: Social media enthusiasts or hobbyists looking for quick and easy video editing.
• Needs: Simplicity, speed, and basic editing features.
• Technical Expertise: Low.

2. Professional Editor

• Description: Content creators, marketers, or video production professionals.
• Needs: Advanced editing options, high-quality output, and customization.
• Technical Expertise: Medium to High.

3. Admin

• Description: System administrator managing the platform's backend and performance.
• Needs: Monitoring tools, error logs, and queue management.
• Technical Expertise: High.

4. Guest

• Description: Potential users exploring the platform's features.
• Needs: Demo access and feature previews.
• Technical Expertise: Low.

5. Visuals Colors and Theme

Unique Color Palette:

• Background: #1E1E2F (Deep Midnight Blue)
• Surface: #2A2A40 (Charcoal Gray)
• Text: #EAEAEA (Soft White)
• Accent: #FF6F61 (Vibrant Coral)
• Muted Tones: #8A8A9D (Muted Lavender Gray)

This palette reflects a modern, professional aesthetic while maintaining a user-friendly and approachable vibe.

6. Signature Design Concept

Interactive Filmstrip Landing Page

The homepage of Divine-Automation will feature an interactive filmstrip design. Users will be greeted with a horizontally scrolling filmstrip that showcases the platform's features as "frames" in a reel. Each frame will represent a key feature (e.g., silence removal, scene cuts, subtitle generation) with animated previews.

Key Features:

• Hover Interaction: Hovering over a frame will expand it, playing a short animation or video preview of the feature in action.
• Dynamic Transitions: Smooth scrolling with parallax effects to create a cinematic experience.
• Call-to-Action (CTA): A glowing "Get Started" button at the end of the filmstrip, leading users to the upload page.
• Background Animation: Subtle, looping animations of film reels and editing tools to enhance the theme.

This design ensures a memorable first impression, engaging users with a visually striking and interactive experience.

7. Non-Functional Requirements

• Performance: The system should process videos of up to 2GB within 10 minutes.
• Scalability: The platform must handle 100 concurrent users without performance degradation.
• Reliability: 99.9% uptime for the backend services.
• Security: All uploaded files must be securely stored and deleted after processing.
• Localization: Support for Indian regional languages in subtitles.

8. Tech Stack

Frontend:

• React (Next.js)

Backend:

• Node.js
• FastAPI (for Python integration)

Database:

• MySQL (primary database)
• Redis (queue management)

AI Models:

• OpenAI Whisper (subtitle generation)
• GPT 5.4 (user-friendly responses)

AI Tools:

• LangChain
• LiteLLM (for LLM routing)

Orchestration:

• Docker
• Kubernetes

9. Assumptions and Constraints

Assumptions:

• Users have access to a stable internet connection.
• FFmpeg and Redis are pre-installed on the server.

Constraints:

• Maximum video file size: 2GB.
• Subtitle generation supports only supported languages by OpenAI Whisper.

10. Glossary

• FFmpeg: A multimedia framework for handling video, audio, and other multimedia files.
• Redis: An in-memory data structure store used for caching and queue management.
• Whisper: An AI model by OpenAI for speech-to-text transcription.
• Parallax Effect: A design technique where background elements move slower than foreground elements, creating a 3D effect.