Training description - AI Vibe Coding Training Program for Non-Developers
Choose your training - AI Vibe Coding Training Program for Non-Developers
- Inter-company or in-company training
- On-site or remote training
General information
This two-day training course enables non-developers to learn how to create a functional web application with the help of artificial intelligence, without limiting the work to a disposable demo or throwaway prototype. It provides the foundations needed to specify an application properly, organize the workspace, understand the essential concepts of software development and guide an AI assistant reliably.
The goal is not to turn participants into professional developers in two days, but to help them understand what they are doing, communicate effectively with AI, verify produced results, version their work and evolve a simple application while staying in control.
Participants leave with a functional web application, a structured Git repository, a specification method, reusable development rules and a clear understanding of key concepts: web application, interface, data, API, terminal, Git, commits, branches, errors, tests and simple deployment.
Objectives and target skills
- Understand the software development basics required for vibe coding: web application, files, dependencies, terminal, browser, data, API and deployment.
- Specify an application properly before asking AI to code: need, users, journeys, data, screens, business rules and acceptance criteria.
- Organize a reliable workspace: folders, Git repository, naming conventions, documentation, secrets management and separation between sources and generated files.
- Use an AI assistant to create, modify, explain, fix and evolve a functional application without blindly executing its proposals.
- Understand and apply essential development rules: version, test, review, document, split changes and protect sensitive data.
- Debug a simple application with AI support by using error messages, logs, the browser console and diagnostic commands.
Audience
Executives, consultants, Product Owners, Product Managers, business managers, trainers, independent professionals, innovation managers and any non-developer wishing to create and evolve a simple application with AI.
Prerequisites
Participants must be comfortable using a computer, a web browser and file management. No prior development experience is required.
Each participant must bring their own computer. An active subscription to an AI assistant suited to code, such as ChatGPT Plus, Claude Pro, Codex or equivalent, is required. A GitHub account is recommended.
Duration
2 days, or 14 hours.
Download the AI Vibe Coding Training Program for Non-Developers
Training content
Understand what will be built
- Difference between idea, prototype, internal application and production application.
- What a web application is: browser, server, interface, data, requests and hosting.
- Frontend, backend, database, API: useful vocabulary without turning the course into a developer specialization.
- What AI can do to code, explain, fix and document.
- Limits of vibe coding: technical debt, security, hallucinations, dependencies and the need for human verification.
Specify before coding
- Describe the business problem and users.
- Define the main journeys and expected screens.
- Identify the data to store, display or modify.
- Formalize business rules and edge cases.
- Write verifiable acceptance criteria.
- Turn a vague idea into a specification usable by AI.
Prepare the workspace
- Install or check the required tools: browser, editor, terminal, AI assistant, Git.
- Organize project folders.
- Understand what a source file, dependency, configuration file and generated file are.
- Create a Git repository and understand its role.
- Define project rules: naming, documentation, structure, test data, secrets and files to exclude.
- Prepare a context file to guide AI.
Understand Git without becoming an expert
- What Git is: repository, history, version, commit, branch and remote repository.
- Why versioning is needed before letting AI modify code.
- Read project state, compare changes and save a step (commit).
- Understand branches as experimentation spaces.
- Avoid dangerous commands and ask for an explanation before execution.
Guide AI to code
- Give AI complete context: objective, constraints, project structure and development rules.
- Ask for a plan before asking for code.
- Generate a first application structure.
- Modify a feature without breaking the existing application.
- Ask AI to explain produced code in simple words.
- Check modified files and reject changes that are not understood.
Build a functional web application
- Create the starter application.
- Add the main screens.
- Manage simple data: list, form, creation, update and deletion.
- Add validations and error messages.
- Improve the user interface with AI.
- Save each important step in Git.
- Document decisions made during construction.
Debug and verify
- Read an error message without panicking.
- Use the browser console and logs.
- Ask AI for a structured diagnosis.
- Reproduce a problem before fixing it.
- Test the main user journeys.
- Check that a fix has not broken another feature.
- Document encountered issues and their solutions.
Evolve the application
- Add a new feature from a user request.
- Split an evolution into small steps.
- Update the specification and acceptance criteria.
- Create an experimentation branch.
- Compare before and after.
- Roll back if an evolution is not satisfactory.
- Prepare a simple release note.
Security, quality and responsibility
- Identify sensitive data and information that must not be sent to AI.
- Understand the notions of secret, API key, environment variable and
.envfile. - Avoid publishing passwords or personal data.
- Distinguish between prototype, internal tool and publishable application.
- Understand risks linked to dependencies and copied code.
- Define a minimum checklist before sharing or publishing.
Simple deployment and delivery
- Understand what deploying an application means.
- Prepare the application for a demonstration.
- Generate a stable version.
- Identify publishing options according to the type of application.
- Write a user README.
- Prepare future evolutions.
Exercises and deliverables
- Complete specification of a simple application.
- Project context file for an AI assistant.
- Project development rules.
- Git repository initialized and used during the training.
- Functional web application built with AI.
- Reusable prompt list for specifying, coding, fixing and evolving.
- Verification checklist before modification, sharing or deployment.
- Application evolution plan after the training.
Access conditions
Registration lead time: from 2 months to 1 week before the training starts, subject to available seats and scheduling constraints.
See the General Terms and Conditions of Sale: https://www.opcoach.com/cgv.
Registration is completed on the OPCoach website using a link provided in the invitation. This registration collects information about the participant, their professional environment, role, expectations and prerequisites before the training starts.
Accessibility for people with disabilities
People with disabilities may have specific accessibility needs to attend the training. Please contact us so that we can discuss suitable arrangements.
Training format
On-site or remote training.
Fees
Inter-company training: EUR 1500 excluding VAT for 2 days.
In-company and custom training: please contact us.
Group size
From 1 to 8 participants, on site or remotely.
Schedule
Schedule available on https://www.opcoach.com/.
Training location
Inter-company training takes place in our InterIndep premises at 2 impasse Michel Labrousse, 31100 Toulouse. The training room is accessible by public transport and by car, and has parking facilities. Several restaurants and a hotel are located nearby. A break area is available for cold and hot drinks during breaks. All these facilities are accessible to people with reduced mobility.
In-company training is preferably delivered at the client’s premises. Hosting conditions are validated before the training.
Remote training is delivered with Zoom and can be organized as inter-company or in-company training.
Teaching approach
Alternating short explanations, guided demonstrations and hands-on exercises on a workstation. The training is organized as a workshop: participants specify, build, test and evolve an application during the course, with support from an AI assistant and under trainer supervision.
Teaching resources
On site: training room with video projector, whiteboard, high-speed Wi-Fi and a training binder provided to the participant.
Remote: high-speed internet access, webcam, free Zoom software, PDF course material sent to the participant. One workstation per participant.
Trainer supervision
Olivier Prouvost, DEA in Computer Science, recognized open source expert and trainer for more than 20 years, developer of AI solutions to optimize SMEs, speaker at international conferences, Eclipse committer on several projects, founder of OPCoach, a Qualiopi-certified training organization since 2020, and ICPF & PSI certified expert trainer and consultant.
Learning assessment
Continuous assessment through the progressive creation of the application: specification quality, repository organization, Git usage, understanding of AI-generated changes, ability to fix an error and add a controlled evolution.
Recognition
A certificate of skills is provided to the participant.
Satisfaction assessment
At the end of the training, the participant completes a satisfaction survey and takes part in a group feedback session. The participant benefits from unlimited email support for practical questions related to the training. After 3 months, or after applying the training in the workplace, a follow-up questionnaire is completed to assess practical use.
Technical, teaching, administrative and financial contact
Olivier Prouvost, +33 6 28 07 65 64, contact@opcoach.com
