Questions to ask the interviewer in data and AI job interviews

Asking the hiring manager

The person you would report to knows the priorities, how success is judged and how much support you would get.

"What would a successful first six months look like in this role?"

Good sign: concrete outcomes — "you'd own the churn model refresh and have shipped one improvement". Worrying: "we'll figure it out", which may mean the role has no clear purpose yet.

"Which recent model or analysis had the most impact on a decision, and how was that impact seen?"

Reveals: whether data science work reaches decisions here. If the manager struggles to name one, the team may produce reports nobody acts on.

"How is work prioritised when several teams want something at once?"

Reveals: whether there is a process or whoever shouts loudest wins.

"How are models or analyses reviewed before they go out?"

Good sign: code review, peer review of methodology, a checklist. Worrying: nothing, especially for a junior hire who will need feedback.

Asking a peer data scientist or engineer

Peers know the daily reality better than anyone else on the panel.

"What does a typical week look like for you — how much is modelling, data preparation, meetings and support?"

Reveals: whether "data scientist" here means building models, writing SQL reports, or maintaining pipelines. None is wrong, but it should match what you want.

"Where does the data come from, and how much time goes into getting it into a usable state?"

Reveals: data maturity. A peer who laughs before answering has told you something.

"How does a model get from a notebook into production here?"

Good sign: a described path — a repository, tests, a deployment process, monitoring. Worrying: "we send the notebook to the engineering team", which often means models stall.

"What's something you've learnt since joining that you didn't expect to?"

Reveals: learning culture and a candid view of the team, often more honest than a direct "what's bad here?".

Asking about AI and LLM work specifically

For GenAI or applied AI roles, these separate teams building real systems from those exploring.

"Which AI features are in production now, and how do you evaluate them?"

Good sign: a test set, human review, metrics tied to users. Worrying: "we check it looks good", which suggests you would inherit untested systems.

"How do you decide whether a problem needs an LLM or a simpler approach?"

Reveals: judgement. A team that uses an LLM for everything may be building expensive, fragile systems.

"What are the data-handling rules for sending company or customer data to model providers?"

Reveals: whether governance exists. Its absence is a risk you would carry.

Asking the recruiter or HR

Recruiters know process and terms, not technical detail.

"What are the remaining stages, and when should I expect to hear back?"

"Is the role new or a replacement? If a replacement, how long was the previous person in it?"

Reveals: growth or turnover. A replacement is not a bad sign on its own; several in quick succession might be.

"What does the role's working pattern look like — office days, hours, travel?"

Ask this here rather than in a technical round, where it can read as your main concern.

Asking a senior leader

In a final round with a director or head of data, ask about direction.

"What are the one or two most important problems data and AI should solve for the business in the next year?"

Reveals: whether leadership sees data as strategic, and whether the role connects to it.

"How is the data team's contribution measured at your level?"

Reveals: how visible and valued the work is.

Questions to avoid

Using the answers

Write down the answers after each round. Patterns across interviewers are informative: if the manager describes a rigorous review process and a peer says nothing is reviewed, you have learnt something important. The last question — "anything you'd like me to clarify?" — also gives you a final chance to fix a weak answer from earlier.

For the rounds these questions close, see the managerial round bank and the ML system design round. Interview preparation, including mock rounds, forms part of career support in the Data Science programme.