Why did you leave (or are leaving) [company]?
Machine Learning / AI
How did you arrive at a specific confidence score threshold?
Are you currently working or not?
Was there any additional training of the model for this task (not LLM)?
Is everything clear to you about the product (media plans), or would you like to clarify something?
How would you now (with current knowledge) solve the classification and routing of requests in a service desk, using a multi-level approach rule-based → ML → LLM fallback → human? Name specific technologies, models, and metrics.
Were you the lead in Data Science in your team at [company]? What approaches and technologies did you originate, and which did the tech lead define?
Compare linear and nonlinear models in classical ML — pros and cons.
What types of tokenization do you know? Pros and cons of each.
Briefly tell me about the team you worked in at [company].
Tell about regularization: what it is, why it is needed, what types exist.
Was it a test/pilot or did it go directly to production and was used by managers?.
What was the product at [company] from the user's perspective (document flow, data extraction from KS-2/KS-3/estimate)?
What is a gradient from a mathematical point of view and how is it applied in machine learning?
Tell me about a classic ML project (not LLM) that you are proud of.
How did you understand whether the model classified the data correctly? What actions were required from the manager?
Tell about yourself, your experience, and key projects (especially about the corporate GPT system at [company]).
How did you measure data extraction quality?