Our services

16h over 2 days

Move from vibe coding to agentic engineering.

A 16-hour training for your developers, tech leads and CTOs: understand agents, master Claude Code, Codex and ChatGPT, and build reliable software with AI — without falling for its pitfalls.

Book a session Book a session

16h

spread over 2 days

3 tools

Claude Code · Codex · ChatGPT

3 formats

on-site · remote · hybrid

Certificate

delivered to each participant

Who it's for

This training is designed for technical profiles who want to understand, use and operationalize agents and AI-assisted development environments:

  • Developers & tech leads
  • Engineering managers
  • VPs of Engineering
  • CTOs and technical leadership

Prerequisites

The training is accessible to anyone with a technical background and a general familiarity with digital tools and development environments.

  • Technical background and comfort with dev environments
  • ChatGPT Business subscription for each participant
  • Claude Max 5x subscription for each participant

LEARNING OBJECTIVES

Three steps to transform your technical teams

Participants progress through a structured journey: understand the fundamentals, master the tools, industrialize the practices.

STEP 1

Understand

Build the foundations to act with clarity.

  • The fundamental principles of agents
  • Prompt and context engineering in a technical setting
  • The impact of model size, context, memory, speed and cost on real use cases
  • The different types of agents, tools, capabilities and orchestration mechanisms

STEP 2

Master

Move from concepts to practice, on real situations.

  • Use Claude Code, Codex and ChatGPT in a guided setting
  • Understand the limits of vibe coding
  • Identify the conditions for moving to an agentic engineering approach

STEP 3

Industrialize

Embed AI into reliable, long-lasting software production.

  • Improve reliability, autonomy and software quality in AI-assisted contexts
  • Apply adapted practices for scoping, refactoring, documentation and testing
  • Set up continuous improvement loops with agents

TOOLS USED

The environments your teams will master

Depending on the sequences and demonstrations, the training relies on three leading tools in the market.

Claude Code

Anthropic's development agent, integrated into the terminal. Used for long work sessions and complex tasks.

Codex

OpenAI's coding agent. Used in parallel to cross-check approaches and understand differences in usage.

ChatGPT

For everyday practical use — synthesis, scoping, exploration — and to contextualize capabilities beyond code.

TEACHING APPROACH

Theory, demos, workshops: from concepts to your real cases.

The training is delivered by a trainer who guides participants through the concepts and their application to concrete cases — ideally drawn from your own stack.

  • Theoretical content

    To understand what really happens under the hood.

  • Live demonstrations

    To see the tools in action on real problems.

  • Discussions & analysis exercises

    To test ideas against your teams' reality.

  • Guided hands-on workshops

    To leave with real experience, not just slides.

LOGISTICS

Everything you need to set up the session

Duration

16 hours over 2 consecutive or spaced days, based on your constraints.

Format & location

On-site, remote or hybrid — based on your context.

Assessment

Ongoing discussions, applied exercises and practical cases.

Credential

A completion certificate is delivered to each participant.

EXPECTED OUTCOMES

What your teams take away

Beyond the learning objectives, the training leaves a concrete trace: mental models, lived experience and best practices ready to use.

  • A structured understanding of the foundations of agentic engineering.
  • Concrete benchmarks on tools and their usage.
  • Hands-on experience working with agents.
  • Best practices to scope, stabilize and industrialize AI-assisted work.

FREQUENTLY ASKED QUESTIONS

You might be wondering

How many participants per session? +

The training is designed to run in-house with a small group. We calibrate the headcount with you upfront to ensure enough hands-on time for everyone.

Do all participants need to be developers? +

No, but all must have a technical background and familiarity with development environments. The training is primarily aimed at developers, tech leads and CTOs.

Which subscriptions are required? +

Each participant needs a ChatGPT Business subscription and a Claude Max 5x subscription. We can help you with the setup if needed.

Can we get the detailed syllabus? +

Yes. The full day-by-day, module-by-module syllabus is shared on request during our first conversation, alongside a proposal tailored to your context.

Can the content be adapted to our stack? +

Yes. The hands-on workshops, and especially the open workshop on day 2, are designed to be grounded in your real cases. We align on this before the session.

BOOK A SESSION

Ready to step into agentic engineering?

Tell us about your team and technical context — we'll come back with a proposal covering format, dates and pricing adapted to your needs.

Request a quote Request a quote hello@insightkeeper.ai