First guided application

How can you move from a simple idea to a testable application in Lovable, without confusing code generation with a reliable result?

  • Lovable
  • Claude Code
  • Codex

1250 € per session, excl. VAT

In short

In a half-day session of 3 hours you build a small application with Lovable based on a fictional requirement. You formulate instructions, examine the results and correct a user journey. You leave with a demonstration prototype, instructions for continuing work on it and a completed test grid.

Why this training?

A convincing interface can conceal an inactive button, an incorrect filter or poorly validated input. The workshop distinguishes the application's appearance from its verifiable operation: every request made to Lovable is linked to an expected behaviour. The first obstacle is often an overly ambitious scope. By limiting the project to a catalogue and a form, participants discover how to clarify an instruction, examine a change and decide what to correct before adding a feature. Running case study: The fictional association « Avelune Loisirs » is preparing its Saturday workshops. It wants an application that allows users to browse six fictional workshops, filter them by category and add a proposal. The prototype has no user accounts, no personal data and no connection to an external service. Transfer to work: Reuse the brief for another fictional internal catalogue, change a single operating rule and rerun the tests. Before any real-world use, have the code, access permissions, data and hosting conditions reviewed by the appropriate people responsible within the organisation.

Objectives of this training

  • Define the scope of a user journey that can be completed during the session.
  • Write an instruction specifying screens, fictional data and expected behaviours.
  • Generate and then modify an application with Lovable under human supervision.
  • Run functional tests and record the discrepancies observed.
  • Document the prototype's limitations and the conditions for continuing work on it.

Who is it for and what are the prerequisites?

Roles and tasks concerned

  • Non-developer staff wishing to explore building a simple application.
  • Team leaders wanting to make an idea tangible before discussing it with a technical contact.

Prior skills

  • No prior AI or programming knowledge required.
  • Ability to use a browser and describe the steps of a simple task.

What you will work on

  1. Scope the workshop catalogue

    Objective: Turn the idea into verifiable instructions. Minimum scope, main user journey, necessary fields and distinction between a demonstration and an operational application.

  2. Obtain a first version

    Objective: Guide generation with explicit instructions. Code generation from instructions; the role of context, constraints and human review.

  3. Correct without expanding the scope

    Objective: Request a targeted change and verify its effect. Reproducible description of a defect, code modification and regression checking. Positioning Claude Code and Codex as other approaches to coding assistance, without mandatory hands-on use.

  4. Test and prepare for further work

    Objective: Establish what actually works. Normal-use tests and invalid inputs; persistence limitations; secrets management and technical review before real-world use.

What you take away

  • Catalogue brief: scope, screens and expected rules.
  • Demonstration prototype built in Lovable, with identified limitations.
  • Log of instructions and corrections made.
  • Completed test grid and handover sheet.

Method and assessment

  • Short explanations followed by immediate practice on the running case study.
  • Individual work followed by work in pairs, with peer review of outputs.
  • Guided group discussion: what works, what needs correcting and what remains a human decision.
  • Preparation: Have a computer and access to Lovable authorised by the organisation; check the necessary access conditions in the official documentation before the session. Prepare the six fictional workshop records and check that no exercise document contains personal data, confidential information or secrets.

Observable criteria: The brief defines the three actions: browse, filter and propose a workshop. ; The filter displays only workshops in the selected category. ; A valid proposal appears in the list during the demonstration session. ; An empty mandatory input produces an understandable message without adding a record. ; Each test states the expected result, observed result and status; remaining defects are flagged..

Frequently asked questions

Do you need to know how to code?

No. The steps are guided; a technical review is still necessary before any real-world use.

Are all three tools used?

Lovable is the basis for the exercise. Claude Code and Codex are positioned as alternative assistance tools; their usage arrangements are covered by their official documentation.

Can internal data be used?

The exercises use only fictional or anonymised cases, in tools authorised by the organisation. No secrets are entered in the instructions.

Will the application be deployed to production?

No. The deliverable is a testable prototype; publication, security and operation require additional validation.

Last updated: