Co-Creator ☞ AI-Powered Video Creation Assistant
An end-to-end data pipeline system that ingests and processes YouTube API data to identify high-performing content patterns. Built NLP-based transformation workflows to analyze script engagement signals and a metadata analytics engine to surface optimization opportunities. Integrated Google Gemini AI and Streamlit to deliver real-time content recommendations including titles, descriptions, and hashtags — translating complex data insights into actionable creator guidance.
🎯Click here and Try it out
(plz get it back up and wait it waking up for a moment)

Project Details
The platform offers three core capabilities:
Automated Scriptwriting: generates tailored content based on video type and key messaging

Script Analysis: provides style recommendations including music selection and visual design suggestions

Video Optimization: analyzes video content, find similar Youtube videos, and creates compelling titles, descriptions, hashtags, and thumbnail designs for maximum audience engagement.

Whether you’re a professional content creator or just starting your video journey, Co-Creator serves as your creative collaborator, transforming ideas into polished, engaging video content through AI-powered insights and suggestions.
Key Features:
- Script generation from basic prompts and video types
- Style and aesthetic recommendations based on target audience analysis
- Complete video optimization including metadata and thumbnail creation
- Gemini AI integration for intelligent content suggestions
Technical Architecture
Data Pipeline & Ingestion
- YouTube Data API v3 for programmatic video metadata ingestion
- OpenCV-based video frame extraction pipeline for visual content analysis
- Automated data collection workflows processing titles, descriptions, thumbnails, and engagement metrics
Transformation & Analytics
- NLP transformation workflows analyzing script content for engagement signals and semantic patterns
- Metadata analytics engine aggregating historical performance data to identify high-performing content patterns
- Feature extraction pipeline combining text, visual, and metadata signals for multi-modal analysis
AI Integration
- Google Gemini API for intelligent content generation and script analysis
- Prompt engineering layer ensuring structured, consistent outputs from LLM responses
- End-to-end workflow orchestration connecting data ingestion → transformation → AI inference → output delivery
Stack Python · Streamlit · Google Gemini API · YouTube Data API · OpenCV · NLP