Ich habe mich auf anderem Weg beworben. Der Vorgang dauerte 3 Wochen. Vorstellungsgespräch absolviert im Juli 2025 bei Sdu (Den Haag)
Vorstellungsgespräch
The job interviews were fairly easy. The first round was conducted by lead data scientists which was more of a vibe check. They only asked for background experience and skills. Then I was invited for the second round that was taken by their AI engineer a data scientist. The job description didn't talk a lot about backend, but the interview involved a couple backend questions and very basic AI questions. They also gave a boiler plate code to finish.
Fragen im Vorstellungsgespräch [1]
Frage 1
Questions on backend: how to deal with a lot of api calls
Very basic questions on rag pipelines, but no in depth questions were asked
Ich habe mich online beworben. Der Vorgang dauerte 3 Wochen. Vorstellungsgespräch absolviert im Jan. 2024 bei Sdu (Den Haag)
Vorstellungsgespräch
Beide gesprekken zijn op locatie (reiskosten worden niet vergoed). Eerst een kennismakingsgesprek met twee technische cases. Opgevolgd door een meer management-achtig gesprek.
Aan de hand van de cases ontdek je dat ze maar wat doen op data science gebied en niet op een solide manier kijken of aanpassingen leiden tot een verbetering van de retriever(of niet).
Bedrijf is onderdeel van een Franse multinational. Het lijkt erop dat het een totale chaos is en dat elk onderdeel van het bedrijf het wiel opnieuw uitvindt.
Fragen im Vorstellungsgespräch [1]
Frage 1
For the purpose of the following discussion topics, assume that we have access to the
following 3 types of documents:
1. Law articles: ground-truth for the law in The Netherlands
2. Case law: past details and rulings from various courts on real cases including
facts and circumstances.
3. Commentaries: interpretations and opinions of experts on particular law
articles, legal concepts and/or jurisprudence
Discussion topic 1: Domain-Specific Re-ranking Model
Outline your approach to developing a re-ranking model customized for the legal
domain. Assume you have access to a high quality varied index of legal documents and
a retriever that returns a set of N such documents.
Example questions:
- How would you go about collecting, processing and annotation of data?
- What types of features/metadata could be particularly useful for re-ranking in a
legal context?
- How would you train the reranker?
- How would you evaluate the reranker?
Discussion topic 2: Improving a RAG pipeline with legal domain knowledge
Our Retrieval Augmented Pipeline has challenges that are particular to the legal
domain. For instance, there are regulations that might be outdated (without this being
clear from the metadata), or a precendent that has been overruled.
Think about more potential failure cases a legal RAG pipeline is especially susceptible
to, and what experiments you would do to solve these problems. If you can’t come up
with extra failure cases, use our examples.