Build your own AI assistant
COURSE
Whereas many lawyers still use AI as a standalone question-and-answer tool, this course teaches you how to use AI as a structural support for your work processes. The premise is simple: if you repeatedly perform the same tasks, make the same analyses, or ask the same questions, you can build a targeted AI assistant that supports this work in a consistent and controlled manner.
LONG
4 hours
LEVEL
Floor
LOCATION
The Hague
Next course: Tuesday, March 31, 2026
, 9:30 a.m. - 2:00 p.m.
€450 excluding VAT
Small group (max. 15)
Hands-on, practical
PO/PE points: 4
(not legal / not tax-related)
About this course
In this course, you will learn how to design and build AI assistants. We will work with concrete examples and live demos to show how different AI assistants function in practice, which design and control choices determine the quality of the output, and where the actual added value lies. Explicit attention is paid to legal prompting as it applies to building and controlling AI assistants, and how this differs from the use of AI in individual chats.
Then you get to work yourself. You develop AI assistants that fit in with your own work and build them step by step, under supervision and with room for questions and refinement.
Next, you will learn how multiple AI assistants can work together within a single work process. We will also show you how to build an AI orchestrator that manages multiple AI assistants.
After completing the course, you will not only have insight, but also concrete, working AI assistants that can be used immediately in your own practice. You will know how to structurally organize, expand, and control AI so that it does not remain an experiment, but becomes a reliable part of your daily legal work.
Practical result
This is a hands-on course. Participants are guided step by step through the process of designing, building, and combining AI assistants.
Afterwards:
Do you know how to build AI assistants (Custom GPTs) in a thorough manner?
you have built at least 3 AI assistants of your own;
you have gained plenty of inspiration for even more AI assistants that are relevant to you;
do you understand how assistants can call each other?
Do you know how an orchestrator works and how to build one?
Program
Introduction: AI assistants for lawyers
We start with the question of what exactly an AI assistant is and when it makes sense to use one in legal work. In doing so, we distinguish between isolated interactions with AI and working with dedicated assistants that fulfill a fixed role and task.
Using concrete examples and live demos, we show how different AI assistants function in practice and what types of work they are suited for. These examples illustrate how AI assistants can support legal processes and where their added value lies.
Designing effective AI assistants
Next, the focus shifts to design. You will learn how to make conscious choices in advance when developing an AI assistant, what steps to take when building an AI assistant, and which tasks are suitable for support by an AI assistant.
Instructing and managing AI assistants
Building an AI assistant requires a different approach to legal prompting than simply asking individual questions. In this section, you will learn how to direct AI assistants in a targeted manner and how to formulate instructions that are appropriate for the purpose.
Building
With this foundation in place, it's time to get started. Participants first brainstorm ideas for AI assistants that fit in with their own work. They then build these assistants, under supervision and with room for questions and refinement.
Collaborative AI assistants
This course component focuses on getting multiple AI assistants to work together. Participants learn how different specialized assistants can be deployed within a single work process, each with their own role and task, so that complex tasks are broken down into manageable steps.
We discuss how AI assistants can complement and follow up on each other, for example by first providing structure, then performing legal analysis, and finally checking or refining the output. The focus is on consciously designing this collaboration, with clear responsibilities for each assistant and maintaining quality and consistency.
Using practical examples, we demonstrate how collaborative AI assistants lead to greater control, better output, and more efficient workflows. This component represents an important step toward building robust AI work processes, in which AI is not used in isolation but functions as a coherent whole.
AI orchestrators
In this course, we focus explicitly on working with AI orchestrators. You will learn what an orchestrator is within ChatGPT, when it adds value, and how it differs from a regular AI assistant.
We demonstrate how to build an orchestrator that does not perform any substantive legal work itself, but directs multiple specialized AI assistants. The orchestrator determines which assistant is deployed, in what order, and under what instructions, while monitoring scope, quality, and consistency.
During the course, we will jointly build a working orchestrator and demonstrate how it can call other Custom GPTs, have their output checked, and combine the results into a single, clear end product. This will teach you how to organize AI structurally within your own legal work processes, rather than using it ad hoc.
For whom
This course is designed for legal professionals, including lawyers, corporate lawyers, judges, public prosecutors, employees of various ministries, employees of enforcement and investigative authorities, and compliance officers who are interested in improving their understanding and skills in applying language models within their field.
This course is also great for those who support legal professionals, such as:
Legal staff
Office directors
Secretaries
Personal assistants
Participants who want to take advantage of advanced technologies to increase their efficiency, improve their services, and implement innovative solutions in their daily operations will benefit from this course.
The program is also suitable for IT professionals within the legal sector who work with legal teams and are involved in implementing and managing AI technologies.
Your teacher
Joyce Boonstra
Former lawyer, Legal Tech Expert & Trainer
Joyce combineert wat zelden samengaat: een juridische achtergrond met diepgaande technische kennis van AI. Ze was een van de eerste juristen in Nederland die taalmodellen structureel integreerde in haar dagelijkse praktijk. Niet als experiment, maar als werkwijze.
While others were still waiting, Joyce was testing. She knows the possibilities and pitfalls of various tools firsthand. Not from theory, but from hours of practical experience.
That combination makes her training sessions different. No slides full of promises, but concrete applications. No hype, but nuance. And always the question: does this work for your practice?
Practical information
DATE AND TIMES
See below.
COSTS
EUR 450, excluding 21% VAT
METHOD OF PARTICIPATION
This course is only offered in person.
LOCATION
60 Koninginnegracht, The Hague
GROUP SIZE
Maximum of 15 participants
PO/PE POINTS
4 points (non-legal/non-tax)
TAKE WITH YOU
Your laptop.
FURTHER REQUIRED
ChatGPT paid account.