Tech_Stack.md

ClipMind AI — Tech Stack

This document outlines the tools, libraries, and frameworks used to build the ClipMind AI platform.

1. Programming Languages & Frameworks

• Backend: Python 3.11 with FastAPI (for high-performance async APIs)
• Frontend: TypeScript, React.js, and Next.js 14 (App Router, Standalone mode for Docker)
• Styling: Tailwind CSS

2. Databases

• PostgreSQL: Primary relational database for user management, roles, classrooms, and video metadata. (Hosted on Render)
• MongoDB: NoSQL document database used for storing large, unstructured data such as generated transcripts, summaries, and key moments. (Hosted on MongoDB Atlas)

3. AI & Machine Learning

Note: To achieve the sub-second processing speeds demonstrated in our evaluation (1.02s per summarization), we utilized the Groq LPU API.

• Speech-to-Text: Whisper model architecture (via Groq for speed)
• NLP Summarization & Intelligence: Large Language Models (LLM) for generating concise summaries, study guides, and quizzes.
• Computer Vision / Video Processing: FFmpeg (for extracting audio tracks from video files prior to transcription).

4. Video Processing & Storage

• Local / Ephemeral Storage: File processing and uploads use the local filesystem during AI inference.
• Tools: ffmpeg-python for parsing video duration and extracting .mp3 audio chunks.

5. UI / UX Libraries

• Framer Motion: For fluid, spring-based animations across the Next.js frontend.
• Recharts: For rendering analytics data and top keywords on the dashboard.
• Lucide React: Iconography.

6. DevOps & Deployment

• Containerization: Docker & Docker Compose (Multi-stage builds for both frontend and backend)
• Cloud Hosting: Render.com (Web Services & Managed Postgres)
• Version Control: Git & GitHub

7. Testing & Evaluation

• Testing Framework: pytest for backend API and pipeline testing.
• Evaluation: Jupyter Notebooks (backend/ai/evaluation.ipynb) for testing AI model latency and summarization quality.