OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
RAG & Knowledge Systems

Launchstack

Launchstack is a TypeScript framework and Next.js reference application for AI-native systems, combining document ingestion, OCR, retrieval-augmented generation, knowledge graphs, LLM abstractions and background jobs.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Launchstack — OpenViST project profile in RAG & Knowledge Systems.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use Launchstack when you want a repository-based TypeScript starting point for document-centered AI systems and can run its web app, database and worker infrastructure.

Who it is for

  • Developers building AI applications that ingest and organize documents.
  • Teams evaluating a self-hosted reference app for document upload and chat.
  • TypeScript developers who want to adapt RAG, OCR or knowledge-graph components.

Documented uses

  • Build an ingestion and OCR pipeline for a searchable document collection.
  • Create a document-based RAG application for querying organized material.
  • Add knowledge-graph-backed retrieval and LLM integrations to an AI application.
  • Run background processing for document ingestion and knowledge-base updates.
Before you choose

Requirements & limits

The repository is public and licensed under Apache-2.0. The engine packages are not yet published on npm; the documented way to try the project is to run the repository.

  • The project is a framework and self-run reference app rather than a hosted document-chat service.
  • The non-Docker setup requires PostgreSQL with pgvector and a separate worker for ingestion.
  • The README documents Docker lite and OCR-enabled stacks; the OCR option adds Docling for Office documents.
  • Chat provider configuration uses CHAT_BASE_URL and CHAT_API_KEY for non-default endpoints; provider-specific variables alone do not configure chat.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Start Launchstack with Docker

Prepare the repository and start its documented lite Docker stack.

Before you start

  • Node.js 20 or later
  • pnpm 10.15.1
  • Docker and Docker Compose
  • A configured .env file containing DATABASE_URL and BETTER_AUTH_SECRET

Steps

  1. Clone the Launchstack repository, enter its directory and install dependencies with pnpm.

  2. Copy .env.example to .env and configure the required database URL and authentication secret.

  3. Run the command below to start the lite stack in detached mode.

    Documented command
    make up-prod

  4. Use the command below to follow logs, the command below to stop while retaining volumes, or the command below to stop and remove volumes.

    Documented command
    make logs
    Documented command
    make down
    Documented command
    make down-clean

Expected outcome

The documented lite Docker stack starts in detached mode; the README estimates about 400 MB RAM for this stack.

Guide limitations

  • The OCR-enabled Docker profile adds Docling for Office documents and has a higher documented memory estimate.
  • This guide covers startup, not application workflows or a verification of successful deployment.
Documentation 1 ↗
Source-based practical guide

Prepare a non-Docker development setup

Configure the database and start the web app alongside the worker required for document ingestion.

Before you start

  • Node.js 20 or later and pnpm 10.15.1
  • PostgreSQL with the pgvector extension
  • A configured .env file with DATABASE_URL and BETTER_AUTH_SECRET

Steps

  1. Clone the repository, install dependencies with pnpm, and configure .env.

  2. Run pnpm --filter @launchstack/web db:migrate to apply both migration sets.

  3. Run pnpm --filter @launchstack/core db:seed if you want the optional sample company, user and document.

  4. Start the web app with pnpm --filter @launchstack/web dev and the worker with pnpm --filter @launchstack/worker dev.

Expected outcome

The README's non-Docker development path starts the web app and its separate ingestion worker.

Guide limitations

  • Stock PostgreSQL without pgvector is insufficient; the migration runner enables the extension and exits with an error if it is unavailable.
  • The web app alone accepts uploads but does not process them; the worker must run alongside it.
  • This guide covers setup and startup, not successful execution of an ingestion workflow.
Documentation 1 ↗
COMPARE OPTIONS

Compare options for Retrieval-augmented applications

These projects document a common task and share a project role. This is a source-based comparison, not an execution benchmark or a guarantee of interchangeability.

ProjectDocumented purposePlatformAccess and requirements
LaunchstackBuild an ingestion and OCR pipeline for a searchable document collection.Cloud, Docker, Self-hosted, ServerThe repository is public and licensed under Apache-2.0. The engine packages are not yet published on npm; the documented way to try the project is to run the repository.
LangsearchCrawl website or local data for a knowledge base.CLI, Self-hosted, Server
Hybrid Search RagCombine document retrieval and language-model generation in a configurable workflow.
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
9 October 2026
Original website title
Deodat-Lawson/LaunchStack · Official homepage title · retained in its source language
How we review sourcesSuggest a correction