Creator Chat

Creator Chat is a self-hosted app for turning public creator content into personalized, source-aware AI chat experiences.

It gives you a full workflow: add a creator, search public content, review and approve the useful signal, build a creator profile, personalize the user experience, and chat with context grounded in the approved corpus.

Creator Chat is not affiliated with, endorsed by, or officially connected to any creator whose public content you ingest. Use it only with content you are allowed to process, and review each source before it becomes memory.

Features

• Multi-creator workspace with separate chat threads.
• Public content collection through Apify-backed platform scrapers.
• Review gate so discovered content is approved before ingestion.
• Source-backed answers with links to the material used.
• Creator persona generation from approved content.
• Creator customization for profile, retrieval mode, colors, and identity.
• User personalization for name, photo, response style, and background context.
• Strict RAG or content-plus-web retrieval modes.
• PostgreSQL and pgvector storage for searchable knowledge.
• Optional background workers for scraping, transcripts, ingestion, and profile jobs.
• Open-source deployment helpers for Render and Vercel.

Stack

• Backend: Python, FastAPI, Pydantic, psycopg 3.
• Frontend: React, Vite, plain CSS.
• Database: PostgreSQL with pgvector.
• AI/providers: Gemini for chat and analysis; OpenAI-compatible APIs can be used for embeddings or transcription paths; Apify is used for social scraping.
• Optional services: Redis-compatible cache, AssemblyAI, Brave Search, Exa.

Prerequisites

• Python 3.11+
• Node.js 20.19+ or 22.12+
• PostgreSQL 12+ with pgvector
• API keys for the providers you enable:
  • GEMINI_API_KEY for creator chat and analysis.
  • APIFY_TOKEN for scraping.
  • OPENAI_API_KEY or EMBEDDING_API_KEY for embeddings.
  • Optional: ASSEMBLYAI_API_KEY, BRAVE_SEARCH_API_KEY, EXA_API_KEY.

Backend Setup

Create and activate a virtual environment, then install dependencies:

cd backend
python -m venv .venv
.\.venv\Scripts\Activate.ps1
pip install -r requirements.txt

Create backend/.env from backend/env.example and fill in the required values:

Copy-Item env.example .env

At minimum, configure database access, JWT_SECRET_KEY, GEMINI_API_KEY, APIFY_TOKEN, and an embedding key.

For local cookies, backend/env.example uses COOKIE_SECURE=false and COOKIE_SAMESITE=lax. Use secure cookie settings for HTTPS deployments.

Create the database and enable pgvector:

CREATE DATABASE creator_chat;
\c creator_chat
CREATE EXTENSION IF NOT EXISTS vector;

Apply SQL migrations from the repository root:

Get-ChildItem backend/migrations/*.sql | Sort-Object Name | ForEach-Object { python run_migration.py $_.Name }

Fresh installs do not create a default account during migration. Create the first user through the app or POST /auth/register after the backend is running.

Run the backend:

$env:PYTHONPATH = (Get-Location).Path
python -m uvicorn backend.app:app --reload --host 127.0.0.1 --port 8000

Optional worker process:

$env:PYTHONPATH = (Get-Location).Path
python -m backend.services.system_worker

Frontend Setup

cd frontend/creator-chat
npm install
Copy-Item .env.example .env
npm run dev

The frontend defaults to http://127.0.0.1:8000. Set VITE_API_BASE_URL in frontend/creator-chat/.env when the backend runs elsewhere.

Key Environment Variables

• DATABASE_URL or DB_HOST, DB_PORT, DB_NAME, DB_USER, DB_PASSWORD
• JWT_SECRET_KEY
• GEMINI_API_KEY
• APIFY_TOKEN
• OPENAI_API_KEY or EMBEDDING_API_KEY
• TRANSCRIPTION_API_KEY or OPENAI_API_KEY
• ASSEMBLYAI_API_KEY
• SEARCH_API_KEY, BRAVE_SEARCH_API_KEY, or EXA_API_KEY
• THREAD_CONTEXT_CACHE_REDIS_URL
• CORS_ORIGINS, FRONTEND_URL, or FRONTEND_URLS
• SCRAPE_MAX_PLATFORMS_PER_SEARCH
• SCRAPE_MAX_ITEMS_PER_PLATFORM
• SCRAPE_MAX_ITEMS_PER_SEARCH
• SCRAPE_MAX_CREATORS
• SCRAPE_MONTHLY_ITEM_ALLOWANCE

See backend/env.example and frontend/creator-chat/.env.example for the full list.

Deployment

render.yaml is a starter Render Blueprint with secrets marked sync: false.

If you use scripts/redeploy-live.ps1, provide Render service IDs through environment variables or flags instead of committing them:

$env:RENDER_BACKEND_SERVICE_ID = "srv-your-backend"
$env:RENDER_WORKER_SERVICE_ID = "srv-your-worker"
.\scripts\redeploy-live.ps1 -Backend -IncludeWorker

The frontend lives in frontend/creator-chat.

The frontend/anti-gravity folder is a small compatibility wrapper for an existing Vercel project whose root directory was already set to that path. New deployments can point directly at frontend/creator-chat and ignore the compatibility wrapper.

Checks

Backend tests:

python -m pytest backend/tests

Frontend lint and build:

cd frontend/creator-chat
npm run lint
npm run build

Open-source readiness checks:

python -m pytest backend/tests/test_open_source_readiness.py

Safety And Responsible Use

• Do not commit .env, API keys, database URLs, local caches, or generated logs.
• Rotate any secrets that were ever committed before publishing a repository.
• Set a strong JWT_SECRET_KEY for every deployed environment.
• Keep COOKIE_SECURE=true when serving over HTTPS.
• Review provider logs and scraped data before publishing demo databases or fixtures.
• Respect platform terms and creator rights. Creator Chat is designed for public content workflows, but you are responsible for how you collect and use content.
• Generated replies are AI outputs based on approved public sources and configuration. They should not be presented as official creator statements.

Contributing

Contributions are welcome. Read CONTRIBUTING.md before opening a pull request.

For vulnerabilities or sensitive issues, read SECURITY.md and avoid posting exploit details publicly.

License

Creator Chat is licensed under the MIT License. See LICENSE.