Legal AI is only as good as the knowledge behind it. At Lefebvre Sdu, we’ve been building that knowledge for 200 years. Day in, day out. Law by law. Word by word. Other legal AI tools run on our data and software. As the Dutch heart of Lefebvre Group, the European leader in legal and tax knowledge, we’re also developing our own platform: GenIA-L, where lawyers and tax professionals don’t just get an answer, but an answer they’re willing to put their name to. More than 75,000 professionals use it every day, accounting for over €100 million ARR across Europe. We’re looking for a Senior AI Engineer to help shape how the intelligent core of that platform evolves.

HOW YOU'LL MOVE US FORWARD

GenIA-L is a research platform for legal professionals. A user asks a question. The platform searches our trusted sources and turns them into a usable answer. This is done with agents: specialised AI components that each take on part of the question and work together. In the legal domain, an answer can’t just sound convincing. It has to be correct, and you need to be able to see which source it comes from. This isn’t a pilot: the platform is in production and used heavily every day.

As a senior, you build alongside the team, but you also keep the bigger picture in view. You know the different parts of the platform and how they affect one another. You help prioritise, spot obstacles early and make sure we’re heading in the right direction, both in how we work and technically. You look ahead: where could we go, and what do we need to know to get there? And you help medior colleagues grow by seeing exactly where they’re stuck, without taking the work off their hands.

You give direction to the agent architecture of GenIA-L: which specialised agents are needed and how they work together.
You break down legal processes and translate them into technical solutions. You decide how the components are best connected.
You make sure our knowledge packages and the pipelines behind them can scale and stay up to date. A knowledge package is an expert-reviewed set of commentary, case law, legislative history and practical guidance for each article of law.
You build the way we measure quality. Using evaluation sets, observability and metrics, you show that answers, source attribution and reliability are genuinely improving. Not based on gut feeling, but on evidence.
You help prioritise the team’s work and flag technical or process risks early.
You explore new developments, test ideas in practice and turn them into proposals that make the product sharper or faster.
You coach medior engineers and work directly with editors and subject-matter experts.

WHAT WE'RE LOOKING FOR

You have a realistic view of LLMs. You see what they can do, but you also know their weak spots. You follow new AI developments closely, but you don’t just follow the herd: you stay critical and aren’t afraid to say so. You can break a complex problem down into smaller pieces and keep oversight of a system with many moving parts. You take ownership, show initiative when things aren’t yet clear and think from the customer’s problem. You enjoy helping others move forward without taking over. And you like working with lawyers and editors, because they know what a good answer looks like.

Extensive experience in software or ML engineering, around 5 years or more. Depth in recent projects counts for more than the number of years.
In-depth knowledge of Python and experience with relevant machine learning libraries.
You’ve built LLM applications and run them in production.
Solid experience with agent frameworks such as LangGraph and LangChain, tool calling, MCP or similar orchestration. Our entire product is built around this. Experience with only a simple search-and-answer setup (RAG) isn’t enough.
You’ve made the quality of LLM systems measurable with evaluation sets and observability, and can give concrete examples. Experience with tools like MLflow and LiteLLM is a plus.
Strong backend experience, including REST APIs and Postgres databases.
Experience with Kubernetes.
Experience in the legal, tax or another knowledge-intensive domain is a plus.

WHAT YOU'RE SIGNING UP FOR

On Tuesdays and Thursdays, the team works together at the office. What colleagues value most here: mutual trust, room to experiment, and a team that helps each other out when someone gets stuck.

A place where you have influence from day one: on the approach, the processes, and the direction.
Colleagues who deliberately chose Lefebvre Sdu over big tech jobs because here, they're building something that doesn't exist yet.
Base salary between €5,500 and €6,450 per month, based on a 36-hour week, since that's full-time for us. Plus a lot of extra allowances (see below).
A bonus tied to company performance: a fixed year-end bonus of 4% and profit sharing of up to 6%.
23 vacation days and a personal choice budget of 12%, of which 8% is holiday pay and 4% is freely allocable. Want more vacation? Just buy it.
A bonus of up to 10% of your salary, based on company performance and individual performance.
A solid pension plan through PGB, including supplementary partner insurance.
Collective insurance plans.
A travel allowance of €0.25 per kilometer or a public transport reimbursement, plus an attractive bike plan. For employees with a commute of less than 10 km or more than 60 km, we have an adjusted arrangement.
Two days a week in the office in The Hague, within walking distance of Voorburg station. The rest of the week, wherever you work best.
Room to grow: in your craft, in your role, in the organization.

Want to know more?

We’d love to tell you all about the team and the role. Send us your CV and you’ll hear from us within 5 working days. Any questions first? Message or call Nadine on +31 6 28556265.

A reference check and a VOG (certificate of good behaviour) are part of the selection process.