rocketride-server

Hey HN, I'm Joshua from RocketRide. We built an open source AI pipeline platform that turns your IDE into an AI command center. The problem: building AI features means weeks of infrastructure plumbing before writing application logic. RocketRide eliminates that with composable pipelines, a unified API, and one-click deploy. The C++ engine handles concurrency, backpressure, and stage isolation at petabyte scale. Pipelines are JSON. 50+ built-in nodes cover OCR, transcription, summarization, redaction, chunking, embedding, plus model integrations (Claude, GPT, Gemini, Grok). MCP server exposes pipelines as tools for Claude/Cursor/Windsurf. Stack: C++ 49.6%, Python 24.7%, TypeScript 17.6%. MIT license. Only prereq: Node.js 18+. Donated by Aparavi after years of enterprise AI deployments. Linux Foundation and Agentic AI Foundation member. Rep

Hey HN! We built RocketRide because we were tired of context-switching between our IDE and browser-based pipeline tools every time we needed to build or debug an AI workflow. RocketRide is a VS Code extension backed by a high-performance C++ engine. You build pipelines visually inside your IDE using a .pipe file format, and the extension renders a canvas builder directly in VS Code. No browser tab. No separate app. Under the hood: - C++ core with native multithreading — built for throughput, not demos - 50+ pipeline nodes: 13 LLM providers, 8 vector databases, OCR, NER, PII anonymization, and more - Multi-agent orchestration with CrewAI and LangChain built in - TypeScript, Python & MCP SDKs to integrate pipelines into your apps - Runs entirely on your own infrastructure (Docker/on-prem) — your data never leaves your machine Qui