Pipeline Stages
Historical Topics
Planned Videos
Book Sources
What It Does
An evolution of the WealthCode pipeline, purpose-built for documentary-style Indian history content. The key differentiator is a historical research sub-system that downloads books from Project Gutenberg and Internet Archive, extracts relevant passages via AI chapter planning, and grounds every claim in actual source material. Features voice cloning with per-scene dramatic intensity, series-aware episode continuity, multi-language podcast generation (English + Telugu), and automated YouTube Shorts extraction.
What Makes This Different
Downloads books, extracts TOC, AI planner picks relevant page ranges, deduplicates across episodes. Every claim traced to book passages with citations.
Chatterbox TTS with 80+ intensity keywords. Battle scenes get dramatic delivery (0.75), narrative scenes stay measured (0.25). Per-scene exaggeration control.
AI decides if a series continues or ends. Previous episode's narration and sources injected into next script. Auto-playlist management on YouTube.
Single script produces: main video + English podcast + Telugu podcast + 3 YouTube Shorts. RSS podcast feed hosted on GitHub Pages via Google Drive.
Tech Stack
Core implementation
Script + research
Voice cloning
Image generation
Video animation
Pipeline state
Backup + podcast hosting
Upload + playlists