Founder here. Back in 2021, I launched ToolJet here on Hacker News as an open-source internal tool builder, connecting data sources (like Google Sheets) to custom UIs via a secure backend. That initial Show HN post surprisingly hit #1 and kickstarted our journey with a wave of early adopters and even our seed funding. Since then, we've grown a lot – adding a workflow engine, a built-in no-code database, tons of integrations – all aiming to make building internal apps faster and easier.
Here’s a quick demo showing ToolJet AI in action: [https://www.youtube.com/watch?v=H8hIfIHMGnE](https://www.youtube.com/watch?v=H8hIfIHMGnE&ab_channel=ToolJ...)
How it works
-Describe your app in plain English:“I need a simple CRM that lets users add contacts, log calls, and see a dashboard of activity.”
-AI generates a Product Requirements Document (PRD): It drafts a spec for your app’s features/screens to make sure we’re on the same page. This even includes data models.
-It builds the app end-to-end: Once you approve the PRD, ToolJet AI automatically creates the UI screens, database tables, queries, and business logic for the app. In a few minutes, you have a working app that you can run and tweak.
- Context-Aware Suggestions: The assistant looks at your app’s current context – it knows which data sources you’ve connected and the widgets on your canvas. This allows it to generate relevant changes (e.g. referencing the right table or component) rather than generic answers.
It’s pretty wild to see an app go from an idea to a prototype without writing code or even dragging components. In early tests, the AI often gets ~80% of the app right. Of course, it’s not perfect – you might still fine-tune the UI or edge-cases – but it saves hours in scaffolding time. Think of it as having a developer who works at superhuman speed (and never gets tired of boilerplate).
Lately, AI code generators (Bolt, Lovable, etc) have gained a lot of attention, promising to replace manual coding by generating full applications from a prompt. While impressive, these tools often create hard-to-maintain code, struggle with state management, and require constant manual fixes as AI-generated code isn't always reliable.
ToolJet takes a different approach: it doesn't generate code—it builds structured, maintainable applications using our low-code framework. If AI doesn't get something right, you can either fix it visually using our drag-and-drop editor or use our AI assistant to guide you step-by-step. The goal is to speed up development without creating a maintenance nightmare.
We’re here, eager to demo, answer questions, and learn from the community. Thanks for reading, and I’m looking forward to an honest discussion.