Platform Overview
Here's a quick rundown of the key concepts in Vulcan. You don't need to memorize all of this — just skim it once so things make sense as you go.
What Vulcan handles for you
Building a web app normally requires an engineering team: someone to set up servers, configure a database, build a login system, manage deployments, handle security, and keep everything running. Vulcan eliminates all of that.
Here's what you get without writing a single line of code or doing any configuration:
- Hosting — your app runs on a cloud platform automatically. No servers to provision or maintain.
- Deployments — every change you make in the chat is built and deployed live within seconds.
- Login & identity — only Veho employees can access your app, and they're automatically identified. You never need to build a login screen.
- Storage — just tell the AI what data you need to store. It sets up the database or file storage for you.
- Pre-built integrations and templates — a growing library of ready-made building blocks covers the most common needs: connecting to external services, adding AI features, charting data, handling file uploads, and more. You drop them in by describing what you want — no wiring required. The library will keep expanding as Vulcan integrates more deeply with Veho's internal systems.
- Scaling — whether one person uses your app or a thousand, it handles it the same way.
- Version history — one click to save a snapshot, one click to roll back if something goes wrong.
The tradeoff is that Vulcan is opinionated — it works best within its guardrails, and there are some things it can't do (see Limitations). But for the vast majority of internal tools, those guardrails are exactly what make it fast.
Research Workspace
Before jumping into the App Builder, it can help to clarify what you're building first. The Research Workspace is a separate space where you have a focused conversation with an AI analyst — not a code generator — to:
- Articulate the problem and who it affects
- Define the data you need and the questions you need to answer
- Surface edge cases and open questions before writing any code
You can get there by clicking Research first on the Vulcan home page, or via the Research link in the project hub nav bar.
When you're done, click Start Building to graduate your research session into a real project. The App Builder receives a full project brief as context — so Claude already understands what you're building on turn one.
Projects
A project is a single app. Each project has its own:
- Code and files
- Chat history with the AI
- Preview and production URLs
- Storage (if you've added any)
- Version history
You can have as many projects as you want. Think of each one as a separate app you've built.
The Chat Interface
The main way you interact with Vulcan is through the chat panel. You describe what you want, and the AI (Claude) writes or updates the code for you. It's a real back-and-forth conversation — you can ask follow-up questions, request changes, or just say "I don't like that, make it simpler."
The AI reads your current code before responding, so it always knows where things stand. It can also fix its own mistakes — if something looks wrong, just describe it and it'll correct course.
→ Tips, sample prompts, and how to handle bad output
Preview vs Production
Every project has two environments:
- Preview — A live URL for testing. This is where your work-in-progress lives. Changes deploy here first. Safe to experiment.
- Production — Your real, final app URL. You explicitly push to production when you're happy with the preview.
Think of preview like a staging area. Break things there freely — production is untouched until you say so.
Versions
Vulcan can save snapshots of your app at any point in time. If you break something, you can roll back to a previous version. You can also name versions (e.g. "before the redesign" or "v1 launch") so they're easy to find.
Get in the habit of saving a version before asking the AI to make big changes. It takes one click and gives you a safety net.
Storage
Your apps can store data. Vulcan supports three types:
- Key-value (KV) — Simple, fast storage. Great for most apps.
- SQL database (D1) — Structured, relational data. Good for complex queries.
- File storage (R2) — For uploading and downloading files and images.
You don't have to worry about setting these up yourself — just tell the AI what you need and it handles the provisioning.
Templates
Templates are pre-built starting points for common app types — CRUD apps, kanban boards, delivery dashboards, and more. Picking a template doesn't lock you in; you can customize everything after.
Components
Components are pre-built building blocks you can drop into any app — connecting to AWS, Google Sheets, adding charts, maps, file uploads, AI features, and more. You don't configure them manually; just describe what you want and the AI wires the right component in.
Background Jobs
Your app can run tasks on a schedule — no configuration required. Just describe what you want:
"Send a daily summary at 9am" or "Sync data from our API every hour"
The AI handles the schedule and the logic. Jobs run automatically on your production app.
Users and Identity
Every app you build on Vulcan is automatically protected by Vulcan's session auth — only logged-in Veho employees can open it. Every request to your app arrives with the user's verified identity already attached.
This means the AI can build features that know who's logged in without any extra setup:
"Show the logged-in user's name in the header" "Only show records assigned to me" "Track who submitted each request"
Every app has a built-in GET /api/me endpoint that returns the current user's email and display name. The AI wires this up automatically — you don't need to configure anything.