NarrateLoop

SYSTEM ONLINE
Autonomous Reddit-to-Video GenAI Pipeline • Live Production Engine
Server Memory
--
Loading telemetry...
CPU Load & Architecture
--
Oracle Cloud Linux VM (systemd)
Daily Ingestion & Render
--
Zero-truncation Shorts (550w)
Next Scheduled Drop
--
YouTube Data API v3 Ingest
Active Video Artifacts & Downloads
TODAY'S BATCH
Loading rendered video artifacts...
Autonomous Pipeline Topology
STAGE 01
Reddit Ingest
OAuth + RSS Proxy
STAGE 02
Contextual NLP
Gender & Voice Map
STAGE 03
Neural Synthesis
Edge-TTS (+30% Rate)
STAGE 04
FFmpeg Filtergraph
Word Sync & Alpha Fade
STAGE 05
YouTube Release
Thumb & Auto Metadata
Technical Implementation Highlights
1. Contextual NLP Gender Classifier
Parses partner markers (e.g. "Husband (41M)" -> Female voice, "Wife (35F)" -> Male voice) eliminating isolated gender misclassification bugs.
regex NLP heuristics • Jenny/Christopher Neural
2. FFmpeg Dynamic Sub-Pixel Compositing
Renders animated word-level ASS subtitles with custom alpha transitions, floating Reddit UI card intro overlays, and strict PTS audio synchronization.
complex_filter • NVENC/x264 • YUVA420p
3. Multi-Tier Fault-Tolerant Ingest
Three-stage fallback architecture: Authenticated Reddit PRAW API -> Direct Datacenter OAuth -> RSS2JSON proxy fallback with regex cleansing.
OAuth2 • RSS2JSON • HTML sanitizer
4. Daemon Supervision & Telemetry
Dual-redundancy Linux systemd service supervision with psutil lockfile coordination, Telegram Bot control, and FastAPI REST endpoints.
systemd • AsyncIO • psutil • FastAPI
Live Production Telemetry & Logs
[--:--:--] Connecting to live telemetry stream...