Guides
- Add login and SSO to your app with Keycloak on Railway
Deploy Keycloak 26 on Railway for login, single sign-on, OpenID Connect and SAML. First admin login, connecting apps, email limits, costs and upgrades. - Run an always-on AI agent on Telegram or Discord
Run Hermes Agent 24/7 on Railway and chat with it on Telegram, Discord, Slack or a browser. Login, persistence, cost estimates, security and OpenClaw. - Database, cache and auth for a vibe-coded app on Railway
Which database, cache and auth server to run on Railway for an app built with an AI tool, and how services connect. A pick-this-if table with licence notes. - Build AI agents visually with Langflow or Flowise
Build AI agents and RAG chatbots on a canvas, without writing a backend. Host Langflow or Flowise on Railway: login steps, estimated costs, licences. - Chat with your company documents on Railway
Build a document Q&A assistant with RAGFlow v0.27 or Onyx v4 on Railway: what each indexes, login steps, estimated RAM and cost, and licence notes. - Shared support inbox and live chat with Chatwoot on Railway
Add a shared support inbox and website chat widget to your app with Chatwoot on Railway: first login, email limits, estimated cost, licence notes. - Deploy n8n on Railway in queue mode with workers
Run n8n on Railway in queue mode: workers, a webhook processor, Python Code nodes and nightly Postgres backups. Steps, costs and upgrades. - E-signatures on Railway: DocuSeal or Documenso
Self-host DocuSeal or Documenso on Railway for document signing without per-envelope fees. Login, e-mail, certificates, costs and licence notes. - Give your AI agent a cloud browser on Railway (CDP)
Deploy a token-protected headless browser on Railway and connect Playwright or Puppeteer over CDP. Steps, measured memory, limits and options. - Headless CMS for your AI-built frontend on Railway
Host Directus 12 on Railway as the backend for a Next.js, Nuxt or Astro frontend: REST and GraphQL APIs, Postgres, file storage, login steps and licence notes. - Headless e-commerce backend for an AI-built storefront
Run Medusa v2 on Railway as a headless commerce backend with admin, worker, search and Postgres, then connect a storefront you built with AI tools. - Run JupyterLab in the cloud on Railway for data science
Run JupyterLab 4 on Railway with pandas and scikit-learn preinstalled, a password, a persistent volume, pip installs that survive upgrades and costs. - Kali Linux in your browser on Railway
Run Kali Linux in a browser tab on Railway for authorised security testing: a KDE desktop or a web terminal, with login, limits, memory and steps. - Linux desktop in your browser on Railway: which to pick
Run a full Linux desktop in a browser tab on Railway. Compare XFCE, LXQt and KDE templates on Ubuntu, Debian, Alpine and Kali, with memory figures and steps. - One API key and spend limits for LLMs on Railway
Put one OpenAI-compatible endpoint in front of OpenAI, Anthropic and other providers with LiteLLM, New API or 9Router on Railway. Steps, limits and licences. - Long-term memory for AI agents with Mem0 or Hindsight
Give an AI agent memory across conversations with Mem0 v2 or Hindsight 1.0 on Railway: API calls, login steps, measured and estimated costs, and limits. - RabbitMQ or Redpanda on Railway: which queue to pick
Compare RabbitMQ and Kafka-compatible Redpanda on Railway: what each is for, login, connecting your apps, costs and licence notes. - Host a Minecraft server with mods on Railway
Run a Minecraft Java server on Railway with Paper, Fabric, Forge or a Modrinth or CurseForge modpack: EULA, TCP address, backups, memory and cost estimates. - MongoDB replica set for Prisma on Railway
Prisma's MongoDB connector needs a replica set. Run a single-node MongoDB 8 replica set on Railway with backups: connection strings, costs, licence. - Logs, metrics and traces on Railway without Datadog
Self-host Grafana LGTM or SigNoz on Railway to collect OpenTelemetry logs, metrics and traces. What each deploys, how to connect your app, and costs. - n8n on Railway: single instance or queue mode?
Start with single-instance n8n and Postgres, or queue mode with workers? How they differ on Railway, what each costs and when to move up. - Self-hosted NotebookLM alternative on Railway: Open Notebook
Deploy Open Notebook on Railway to chat with your own PDFs, web pages and videos. Login, AI providers, costs and licences, plus DeepWiki for code. - Self-hosted Notion and Miro alternative: AFFiNE on Railway
Run AFFiNE 0.27 on Railway for docs, whiteboards and databases: admin login, adding people, bring-your-own-key AI, seat licensing and costs. - Odoo or ERPNext on Railway: which open-source ERP to run
Compare Odoo 20 Community and ERPNext v16 on Railway: services, login, estimated cost, licences, scaling limits and what is not included. - Postgres backups and point-in-time recovery on Railway
Back up Postgres on Railway and restore to a chosen time. Compare a PITR template with WAL archiving to Databasus scheduled backups, with steps and costs. - Privacy-friendly analytics on Railway: Plausible or Umami
Replace Google Analytics with self-hosted Plausible or Umami on Railway. Setup steps, first login, email, costs and licence notes for both. - Run a private ChatGPT-style chat on Railway
Deploy Open WebUI with Ollama, SearXNG and pgvector on Railway, or LibreChat with MongoDB. Login, costs, model limits and licence notes. - Run Claude Code in the cloud on Railway
Run Claude Code, Codex or OpenCode on an always-on Railway dev box, so agents keep working when your laptop is closed. Steps, costs and options. - Where to keep API keys instead of .env files: Infisical
Self-host Infisical on Railway to store API keys per project and environment, sync them to apps and CI, and keep an audit log. Steps and notes. - Self-host Dify on Railway for LLM apps and RAG chatbots
Deploy Dify 1.17 on Railway: console, API, workers, Postgres, Redis and Weaviate behind one URL. Login steps, costs, upgrades and licence notes. - Self-host GitLab with CI on Railway: what it needs
Run GitLab CE 19 on Railway with SSH access and a shell CI runner. What the template deploys, memory needs, cost estimate and CI limits. - Self-host Langfuse on Railway for LLM tracing and cost
Run Langfuse v4 on Railway to trace LLM calls, track token cost, manage prompts and run evals without sending data to another SaaS. Steps and notes. - Self-host Metabase on Railway for team BI dashboards
Give your team dashboards on your app's database with Metabase on Railway: login, memory needs, email limits, AGPL licence note, upgrades and Superset. - Self-host Nextcloud on Railway as a Drive or Dropbox alternative
Run your own Google Drive or Dropbox alternative with Nextcloud 34 on Railway: login, volume vs Bucket storage, email limits, AGPL note and upgrades. - Self-host Plane as a Jira or Linear alternative on Railway
Deploy Plane, an open-source Jira and Linear alternative, on Railway in one click: first login, 12 services, estimated cost, AGPL note and upgrades. - Self-host PostHog on Railway: what it needs and costs
Run PostHog's self-hosted stack on Railway: 23 services, first login, memory needs, costs, upgrades and licence notes before you deploy. - Self-host Supabase on Railway for a Lovable or Bolt app
Run your own Supabase on Railway: Postgres, Auth, Storage, Realtime and Studio in one stack. Login steps, API keys, costs, upgrades and limits. - Self-host Twenty CRM on Railway as a HubSpot alternative
Run Twenty, an open-source CRM, in your own Railway project: first login, Gmail and Outlook sync, estimated cost, AGPL licence note and upgrades. - Host SillyTavern online on Railway
Deploy SillyTavern 1.19.0 on Railway with a login and persistent storage so your characters and chats work on any device. Setup, costs, security. - Schedule social media posts with AI: Postiz on Railway
Self-host Postiz v2 on Railway to schedule posts across social networks: Temporal included, first login, network setup, costs and licences. - Vector database for RAG on Railway: pgvector
Which vector database for RAG? Use pgvector if your data is already in Postgres. Deploy Postgres 18 with pgvector, pooling and backups on Railway. - Web scraping API for LLM apps: Firecrawl and Crawl4AI
Turn websites into clean markdown for your LLM app with self-hosted Firecrawl or Crawl4AI on Railway: steps, cost estimates, licences and limits. - Self-hosted WhatsApp API on Railway with Evolution API
Send and receive WhatsApp messages from your app with Evolution API on Railway: setup steps, ban risk, licence conditions, memory figures and upgrades. - Host a fast WordPress site on Railway with Redis
Host WordPress on Railway with MariaDB, a Redis object cache, tuned PHP and a real cron schedule. Deploy steps, login, custom domain and cost estimates. - Airflow, Temporal or Windmill for background jobs
Compare Apache Airflow 3, Temporal and Windmill as workflow engines on Railway: what each runs, how to log in, estimated costs and licence notes.