Start typing to search pages and blog articles.

All training courses
Build & apply

AI Accelerator for engineers: build agentic AI

Build a working AI agent with your own engineering team. Modular programme in the Benelux: coding agents, agentic AI and spec-driven development.

Format
Modular programme (2 to 4 weeks, tailored) or Guided sessions, labs and a build sprint
Level
Hands-on
Audience
Software engineers and technical professionals
Location
In-house or online
Language
English or Dutch

Trusted by, among others

KPN
Schiphol
Rabobank
TNO
Amsterdam UMC
Loyens & Loeff
NautaDutilh
Sdu
Tilburg University
Universiteit Leiden
Radboudumc
LUMC
OLVG
Aon
Grant Thornton
Gemeente Den Haag
Provincie Fryslân
KPN
Schiphol
Rabobank
TNO
Amsterdam UMC
Loyens & Loeff
NautaDutilh
Sdu
Tilburg University
Universiteit Leiden
Radboudumc
LUMC
OLVG
Aon
Grant Thornton
Gemeente Den Haag
Provincie Fryslân

The AI Accelerator is our flagship programme for engineers who want to work with AI-native development, coding agents and agentic AI. It is modular: instructor-led sessions, hands-on labs and a joint build sprint in which every team delivers a working agentic AI application. We tailor the length and depth to your team.

About the AI Accelerator

The AI Accelerator takes your engineers from loose experiments to a working solution. You work in small teams. Each team picks its own use case and builds it out, step by step, into an application you can actually put to use afterwards. The programme is modular: together we decide how many weeks and how much depth fit your team.

The approach is practical. Each session covers one technique. In the lab that follows, you apply it straight away to your own code. Theory stays short and most of the time goes to building.

The build sprint is the final piece. Your team delivers an agentic AI application with tests, guardrails and documentation, and presents it in a live demo. What you build connects to the work of your own teams or your clients.

Where the AI Accelerator fits in the learning path

All our programmes can be taken on their own or as a learning path. The Accelerator is the technical side of building and applying AI. If you want to lay the groundwork first, start with AI literacy or Advanced AI training. Teams that want to embed AI in their day-to-day way of working should look at AI Way of Working.

For non-technical teams, an AI Hackathon is a better fit: a single day instead of a multi-week programme. Want to train large numbers of engineers, or give teams extra support during and after the build sprint? See our AI programme at scale. Want to introduce AI more broadly than training alone? See our AI consulting for organisations.

Want a promising idea built or integrated afterwards? Then we work with our sister company NewWorks, which handles the technical development.

Good to know

The Accelerator requires an active software engineering background. You write and debug Python on your own and are comfortable with Git, GitHub, LLM APIs and the relevant AI frameworks. The programme runs in English or Dutch, on-site anywhere in the Netherlands, Belgium and Luxembourg. A group usually has 6 to 30 participants. With more than 15 participants we bring two trainers.

Outcomes

What you will learn

  • You can scope, build, deliver and document an agentic AI solution on your own.
  • You direct coding agents such as GitHub Copilot, Cursor or Claude Code at an advanced level.
  • You work spec-driven: from specification to implementation and delivery.
  • You apply agentic patterns: tools, retrieval, guardrails, evaluation and deployment.
  • You deliver a working, production-grade agentic AI application as a team.
Programme

How the programme is structured

modular, 2 to 4 weeks, tailored

  1. Module 1: agent foundations

    module 1

    Getting to grips with coding agents, first agentic building blocks

    • Architecture and inner workings of coding agents
    • Prompting, reference systems and configuration files
    • Code generation, testing, debugging and Git
    • MCP architecture and tool integrations
  2. Module 2: from specification to system

    module 2

    Spec-driven development and agent design

    • The spec-first approach and project set-up
    • From specification to implementation
    • Agent design, structured outputs and tools
    • Retrieval augmented generation and dependency injection
  3. Module 3: build sprint

    module 3

    Build, harden, document and present

    • Evaluation, guardrails and orchestration
    • Multi-agent patterns, testing and observability
    • Deploying the application
    • Live demos and a joint retrospective
Applications

Where AI helps

Internal developer copilot

An assistant that helps your engineers with code, reviews and documentation inside your own codebase.

RAG assistant on your documentation

An agent that answers questions based on your internal knowledge, manuals and tickets.

Customer-facing AI agent

An agent that handles customer questions, takes actions and escalates to a colleague in time.

Automation in the delivery pipeline

Agents that take over tasks in CI/CD, testing and releases, so your teams ship faster.

Request in-house training

Training tailored to your organisation

We are happy to build a programme that fits your practice, team and learning goals. Fill in the form and we will schedule an introductory call within one working day.

  • Fully tailored

    Content and examples matched to your sector, team and way of working.

  • At your offices or online

    4 or 8 hours, groups of up to 20 participants, in English or Dutch. Anywhere in the Benelux.

  • Proof of attendance

    Attendance is registered; for legal training, CPD points may be possible.

We usually reply within one working day. We only use your details to answer your request.

Deliverables

What you receive

Live guided sessions with technical explanation
Hands-on labs throughout the programme
Joint build sprint with a live final demo
Certificate of participation
Track record

Training in action

From in-house workshops to keynotes for hundreds of participants

Frequently asked questions

What prior knowledge do I need?

An active software engineering background. You write and debug Python on your own and are comfortable with Git, GitHub, LLM APIs and the relevant AI frameworks.

How long does the AI Accelerator take?

The programme is modular: usually 2 to 4 weeks, depending on the depth you want. We shape it around your team and timeline, with instructor-led sessions, hands-on labs and a joint build sprint.

Do we deliver something that works?

Yes. In the build sprint every team builds a production-grade agentic AI application, with a live final demo and a joint retrospective.

How many participants can take part?

A group usually has 6 to 30 participants. With more than 15 participants we bring two trainers, so every team gets enough support.

Can we add extra support?

Yes. Mentoring during the build sprint, assessment and office hours are available as add-on services. See our [AI programme at scale](/ai-programme-at-scale/).

Ready for this training?

Request an in-house training, or start with a no-obligation introductory call.

Prefer email? secretariaat@adviesentraining.ai

AI Quickscan

Book an introductory call

Loading calendar…