#107 · Primary category: Knowledge Base & RAG

LaunchStack

claude docker knowlege-graph langchain mcp n8n neo4j nextjs ollama open-source openai pgvector postgresql rag self-hosted typescript workflow-automation

AI-powered StartUp Accelerator Engine built with Next.js, LangChain, PostgreSQL + pgvector. Upload, organize, and chat with documents. Includes predictive missing-document detection, role-based workflows, and page-level insight extraction.

Project last updated:08/30/26

GitHub Stars

886

Forks

123

Contributors

35

License

Apache-2.0

Why we included this project

LaunchStack is a self-hosted starting point for anyone who has to move a steady stream of application and diligence documents through review, which is most teams running a startup accelerator or incubator. Uploaded files, with optional OCR for office formats, get embedded into a pgvector-backed store, and a chat interface answers questions against that corpus rather than from memory. On top of plain document Q&A it flags documents that look missing from a submission and extracts insights page by page, so reviewers can chase any claim back to its source. The code is a TypeScript engine with ingestion, retrieval, and chat split into reusable ports, wrapped in a Next.js reference app you can deploy as-is. One practical note: the engine packages are not published to npm yet, so for now you consume the project by running the repository itself.

Articles for this project

No articles for this project yet.

To suggest a topic or contribute an article, contact us.

Related projects in this category