FYF Video Pipeline — Autonomous Brand Video Studio
Verified outcome — The 20 QA gates I defined decide what ships — the pipeline rejects its own output rather than publishing it. Built on Google Cloud Run; Google Cloud Agentic Cinema Hackathon (Devpost ClickHouse Partner Track) entry.
- My verified role
- Lead Architect & AI Agent Director / Builder
- Classification
- Feature
- Primary tools
- Next.js, Remotion, FastAPI
- Public evidence
- Demo videoSource code
Source codeThe Problem
Producing localized high-retention video ads and explainers traditionally requires expensive video crews, disjointed editing tools, and manual telemetry tracking.
My Role
Lead Architect & AI Agent Director / Builder
What I Personally Directed
Autonomous creation-loop architecture, brand & scene immutable locking, Remotion rendering pipeline, Google ADK & Gemini TTS orchestration, ClickHouse telemetry outbox, and Model Context Protocol (MCP) Data Officer integration.
The Solution
A full-stack autonomous brand video studio that orchestrates story generation via Google ADK & Gemini, deterministic multi-format video rendering via Remotion, sanitized telemetry ingestion into ClickHouse Cloud, and natural language analytics via an MCP Data Officer.
Verified Outcome
The 20 QA gates I defined decide what ships — the pipeline rejects its own output rather than publishing it. Built on Google Cloud Run; Google Cloud Agentic Cinema Hackathon (Devpost ClickHouse Partner Track) entry.