RocketRide

RocketRide is an AI pipeline builder and runtime for developers creating document processing, search, and agent workflows. Its visual editor sits in VS Code, alongside the application code that uses those workflows.

• Visual workflows: Compose pipelines on a canvas and save them as portable JSON files that you can review and version in Git.
• Document processing and search: Combine OCR, personal-data anonymization, embeddings, and vector search; the included document processor extracts and cleans text using local models, and Qdrant can run locally.
• Agent workflows: CrewAI and LangChain support agent coordination and memory shared across pipeline runs.
• Application integration: Python and TypeScript SDKs connect pipelines to apps, MCP exposes them as tools for AI assistants, and Python extensions add custom nodes.
• Execution visibility: Inspect model calls, token usage, latency, and pipeline execution from the editor to diagnose workflow behavior.

The engine is open source under MIT and runs on your own machine, in Docker, or on an on-premises server. Local operation doesn't require an account. RocketRide Cloud hosts the engine and accepts the same pipeline files. OpenAI and Anthropic nodes use external services. The SDK can fall back to RocketRide Cloud when it doesn't find local engine connection details.