A purchasing approval app
Generate an intake form and review queue over a test database. Explicitly separate the requester from the approver and constrain the write queries.
Generate internal app components, queries and data bindings with AI, then refine them visually. ToolJet offers cloud and self-hosted workflows, with exports that still require its application runtime.
Shortlist ToolJet for internal business software when prompt-generated scaffolding and visual query-based refinement fit the team. Check the data actions and the runtime plan before assuming the generated configuration is ready for employees.
ToolJet’s current AI documentation describes application generation and changes to components, queries and data bindings from prompts. It also documents an MCP path for building ToolJet apps from an external coding agent. Both routes produce a ToolJet application, so the distinction to review is where prompting happens and which account pays for the AI, not whether the runtime has disappeared.
This is a useful shortlist for an internal data workflow with tables, forms and controlled actions. A builder can start from generated scaffolding and refine the application visually. Exporting a ToolJet application produces JSON configuration that another ToolJet workspace can import, optionally including table schemas. Keep that separate from a conventional source-code export, and test the actual datasource and permission setup on the chosen cloud or self-hosted instance.
Suggested briefs to evaluate the workflow, based on documented capabilities.
Generate an intake form and review queue over a test database. Explicitly separate the requester from the approver and constrain the write queries.
Provide a filtered record view and a limited update form. Review generated SQL and parameters before allowing the interface to modify operational data.
Give the generator a sample schema and list the allowed read/write actions for each user type.
Use the built-in prompt builder or an eligible external-agent/MCP workflow, accounting for the different model billing.
Review generated queries and component bindings, and use visual editing for exact layout and workflow changes.
Verify the live app with end-user access, then export/import the configuration if cross-instance continuity matters.
Each fact includes its scope and a link to the official documentation.
ToolJet documents prompt generation of app components, queries and data bindings, with further AI edits in the builder.
ScopeToolJet applications, including an external-agent entry point through MCP.
ToolJet provides cloud use and documented self-hosted deployment. The app still needs its platform runtime and configured data services.
ScopeHosting model and licensed capabilities should be checked together.
Apps export as JSON for import into another ToolJet workspace, with an option to include related database table schemas.
ScopeApplication configuration and schemas, not an independent generated frontend/backend repository.
Review builders, end users, app limits, AI credits and self-hosted entitlements. External-agent model usage has a different billing path from built-in ToolJet AI.
ScopeCurrent plan dimensions; no comparable numerical total is asserted.
Create an expense-approval tool where employees submit their own expenses and a manager approves them through a reviewed database query.
This is a suggested evaluation task, not a test result. SiteKeen has not completed hands-on testing of ToolJet. Our recommendation combines official documentation with editorial judgment. How we review →
Compare 20 AI app-building workflows, from generated web apps to mobile and internal tools. Check AI scope, data, source access, runtime and maintenance.
Explore the picksCompare AI internal tool builders for portals, operations and database workflows. Check permissions, runtime dependencies and the maintenance handoff.
Explore the picks