Jobs › Argonne National Laboratory
Scientific Data Services Engineer – AI & HPC
Posted Apr 23, 2026 · We last checked this listing on Oct 6, 2026
Apply at Argonne National Laboratory
Likely interview questions for this role
Written from this job description, not a generic list. Each one notes what the interviewer is really checking.
About your past
When you start with a new science team, how do you learn enough about their data to decide how it should be structured for search?
What they are checking: Whether they listen to domain scientists and let the science drive the data model instead of forcing a generic schema.
Tell me about a time a data pipeline you built broke in production. How did you find the cause?
What they are checking: Whether they debug systematically and build in monitoring, since silent data errors can ruin research results.
Tell me about a project where you worked closely with scientists. What did you learn from how they think about data?
What they are checking: Whether they thrive in the highly collaborative, cross-discipline environment this group depends on.
What would you do
A model team wants to train on a dataset that lives across several file systems in formats like HDF5 and NetCDF. How would you make it easy for them to stream it?
What they are checking: Whether they know how scientific file formats and parallel file systems behave when feeding AI workloads.
Two science projects want different features from your data service and you only have time for one. How do you decide?
What they are checking: Whether they can weigh scientific impact and number of users and explain a fair decision to both groups.
About the work
How would you design a system to ingest and index a few petabytes of mixed scientific data so researchers can search it quickly?
What they are checking: Whether they can think about ingestion, metadata, and indexing at real scale rather than at laptop scale.
What search technologies have you used, such as Elasticsearch, OpenSearch or vector databases, and how did you choose between them?
What they are checking: Whether they understand the tradeoffs between keyword, faceted and vector search for different kinds of scientific queries.
How would you expose the same dataset both to people through a web app and to AI models through a high-performance API?
What they are checking: Whether they can serve two very different consumers without duplicating data or creating two systems that drift apart.
How do you think about provenance and versioning for scientific data that keeps changing as new runs come in?
What they are checking: Whether they understand that reproducibility in science depends on knowing exactly which data a result came from.
What makes a scientific data web application actually usable for a researcher who is not a software person?
What they are checking: Whether they care about the end user experience and not just the backend architecture.
How have you used HPC resources, schedulers or workflow tools in past data projects?
What they are checking: Whether they can work inside a supercomputing environment where jobs, storage and networks behave differently from the cloud.
Practice this interview out loud.
Offer builds a real interview for this exact role at Argonne National Laboratory from your resume and this job description, asks the questions one at a time, and tells you what landed. The first one is free.
Practice this out loudThe full job description
As published by Argonne National Laboratory.