Interview Practice Questions

Describe the lifecycle of an ML task in your team, from problem definition to deployment.

How does your team define and assign tasks?

Tell us about yourself and the project you consider most interesting technically, in English.

What is your experience with classical ML? How familiar are you with tabular data and non-neural approaches?

You wanted to use ML to influence the business. Did you achieve that? How do you measure business impact?

Give an example of an unsuccessful ML hypothesis: what did you try, why did it fail, and how did you resolve the problem?

How is work organized in your team? Who defines tasks, how do you break them down, and how do you plan sprints?

Describe your biggest failure at work. What went wrong, and what did you learn?

How do you break down and estimate ML tasks? How is your team's roadmap organized?

Describe the most challenging project you have worked on.

What do you do when the data is insufficient or unclear?

Describe the most interesting and challenging projects in your experience.

Describe your team's structure and how responsibilities are distributed.

How does your team move from a business problem to an ML solution?

Describe one of your projects in detail: what was the problem, what was your role, and what results did you achieve?

What was the most difficult part of recommending matching clothes?

Give an example of a project that did not succeed. How did you decide to shut it down?

What did you learn from your projects? How do you deal with challenges when a model doesn't work?

How would you persuade a skeptical team to put your work on its backlog?

У вас был ли опыт мониторинга моделей на продакшене с целью того, чтобы понять, нужно ли делать рекалибровку моделей, пристроение моделей?