Streamlit
PythonNVIDIA API
SQLite
Plotly
Gemma 4
YouTube Pattern Finder is an automated AI intelligence dashboard built to reverse-engineer high-performing video strategies from the YouTube attention economy. Instead of manually analyzing trending content, it leverages a multi-agent orchestration pipeline. By inputting a channel URL or keyword, the system pulls video metadata, renders thumbnails in a fluid grid, and runs a sequential multi-agent processor (using NVIDIA Integrate API and GPT-4o) to analyze hooks, thumbnail visual cues, and title mechanics. Historical analyses are cached in SQLite, and trends are visualized using Plotly Express.
Features
- •Multi-Agent Analysis Pipeline: Orchestrates Hook, Thumbnail, and Title Agents to deconstruct narrative structure, visual cues, and psychological leverage.
- •Channel & Niche Analyzer: Dynamically pulls video metadata and renders fluid grids for rapid visual and performance pattern identification.
- •Pattern Library & Trends: Caches historical analyses in a thread-safe SQLite database, displaying interactive Plotly Express data visualizations.
- •Automated Keep-Alive Workflow: Configures a GitHub Actions cron job to periodically ping the Streamlit app, preventing it from spinning down.