Rose engineering

Build, run, and extend Rose's AI-first inbound marketing platform.

Set up your workspace Explore the architecture

Find your path

• Product agents

  Understand the agents that engage, qualify, and convert B2B buyers.

  Explore Rose agents →

  Open the Website Agent docs →

• Backend

  Work with Rose APIs, orchestration, retrieval, ingestion, and shared services.

  Backend overview →

  Set up the backend →

• Frontend

  Build the widget, backoffice, playground, browser extension, and shared UI.

  Frontend overview →

  Set up the frontend →

• Platform

  Follow data through Rose and operate its configuration and observability systems.

  Platform architecture →

  Observability →

Platform at a glance

Rose connects buyer-facing React experiences to Cloudflare edge services and FastAPI applications. LangGraph agents coordinate model calls and retrieval across LightRAG, Neo4j, MongoDB, Redis, and Supabase.

See the system architecture →

Local setup

You need Python 3.12+, Node.js 24 LTS, Docker, Google Cloud SDK, Terraform, Poetry, and just.

Create an isolated worktree for each change:

git-add-worktree feature/my-branch

Then bootstrap the workspace:

./bootstrap.py

Use ./bootstrap.py --skip-env when you do not need environment files, or ./bootstrap.py --force to recreate the Python environment.

Read the development workflow →

Daily work

• Develop and test the backend
• Develop and test the frontend
• Work with Rose configuration
• Use Rose development tools
• Write and publish documentation