CreatorOS — Autonomous Content Pipeline & Postiz Engine

Turn 1 raw video into 12 platform-native social assets. CreatorOS uses Groq AI & Postiz API to transcribe, generate, review, and auto-schedule posts across Bluesky, LinkedIn & Instagram

Comment

Inspiration

Content creation in 2026 is no longer just about editing a single great video—it is about distribution velocity. Creators, startups, and developer advocates spend up to 80% of their workflow time manually re-editing transcripts, writing platform-specific hooks for LinkedIn, Instagram, and Bluesky, generating promotional thumbnails, and scheduling posts across disparate dashboards.

We were inspired to build CreatorOS to eliminate this friction entirely: What if you could drop a single raw video into a pipeline and get an entire automated multi-platform campaign executed in seconds?

The Problem

  1. Repurposing Bottleneck: Converting a 5-minute video into platform-native text, carousels, and hooks requires hours of repetitive copy-paste work.
  2. Context Blindness: Generic AI writing tools generate robotic text that ignores platform norms (e.g., LinkedIn professional tone vs. Bluesky concise tech banter vs. Instagram visual CTAs).

3. Disconnected Tools: AI generators don't schedule posts, and social schedulers don't generate content—forcing creators to juggle 5 different tabs.

What We Built

CreatorOS is an end-to-end autonomous content pipeline and social media execution engine. It accepts raw video files (MP4/MOV) or topic ideas, transcribes speech with Groq Whisper V3, extracts actionable insights using Groq Llama-3.3-70B, generates 12 platform-native social content assets, provides an interactive review suite with a custom thumbnail generator, and automates distribution via the Postiz API.

All of this is wrapped in a vibrant, high-energy Neo-Brutalist design system featuring an interactive React Flow visual execution graph.

How It Works

  1. Source Ingestion: Upload a video or enter a topic prompt on the Create Campaign stepper interface.
  2. AI Transcription & Extraction: Audio is processed via Groq Whisper V3, followed by Groq Llama 3.3 70B JSON content extraction for summaries, tone, viral hooks, and hashtags.
  3. 12-Asset Generation: Generates multi-platform posts tailored specifically for Instagram, LinkedIn, and Bluesky.
  4. Interactive Review & Customization: Creators review content cards, make inline text edits, and utilize the built-in ThumbnailCreator for post visuals.

5. Postiz Automated Queueing: 1-click approval dispatches posts directly to connected social accounts via Postiz API integration, updating live status on our interactive workflow graph.

Key Features

  • ⚡ Neo-Brutalist Visual Identity: Standout high-contrast aesthetic with heavy borders, vibrant palette, and crisp typography.
  • 🎙️ Groq Whisper Transcription: Sub-second speech-to-text conversion.
  • 🧠 Llama 3.3 70B Intelligence Engine: Contextual analysis that understands platform-specific nuances.
  • 📊 React Flow Visual Pipeline: Real-time visual node graph (@xyflow/react) displaying campaign processing states.
  • 🎨 Integrated Thumbnail Creator: Drag-and-drop banner creation right within the campaign workspace.

* 📅 Postiz Multi-Channel Queue: Unified social calendar for tracking queued and published posts across networks.

Technical Implementation

CreatorOS is constructed with a modern, production-ready stack:

  • Frontend: Next.js 16 (App Router), React 19, TypeScript, Tailwind CSS 4, Framer Motion, Lucide Icons.
  • Pipeline Visualization: @xyflow/react (React Flow v12) render graph displaying node execution states (upload ➔ transcribe ➔ generate ➔ postiz_publish).
  • Database & Persistence: Supabase PostgreSQL with @supabase/ssr and database migration schemas for campaigns, media, and scheduled posts.
  • AI Infrastructure: Groq API SDK executing llama-3.3-70b-versatile and Whisper V3 with JSON format enforcement.

* Publishing Architecture: Custom PostizProvider adapter integrating directly with Postiz public API v1 endpoints.

Challenges We Faced

  • Strict JSON Response Enforcement: Ensuring LLM outputs consistently return fully structured platform object arrays without format breaks. We resolved this using Groq's response_format: { type: 'json_object' } combined with robust Zod/TypeScript schema defaults.
  • State Sync with Visual Graphs: Synchronizing API progress with real-time React Flow graph animations required carefully managed React state hooks and event triggers.

* Seamless API Fallbacks: Building automatic mock provider fallbacks ensured that judges can experience the full interactive app even if third-party API rate limits occur during evaluation.

What We Learned

  • High-speed inference (like Groq LLM processing) transforms AI content creation from a slow waiting game into an instantaneous background pipeline.

* Pairing visual execution diagrams with automated backend workflows dramatically improves creator trust and control over AI outputs.

Why This Matters

CreatorOS empowers solo creators and lean growth teams to compete with enterprise media agencies by accelerating content distribution velocity by 10x while drastically cutting manual effort.

What's Next

  • Automated Video Clipping: Auto-generating 60-second vertical video clips from long-form content.
  • Direct Postiz Webhook Listener: Real-time analytics tracking for views, likes, and shares directly inside the CreatorOS analytics view.
  • Multi-Language Localization: Translating video transcripts and generating localized campaigns for global audiences.

Built With

Share this project:

Updates

Submission history