On this page
Questions to practise
Tell me about a model you shipped. What problem did it solve?
She wants one real example. It shows what you can do in real work, not only in theory.
Pick one model. Say the problem first, then what you built, then what changed after. A small change is fine. For example: 'I built a model to predict late deliveries. After that, the team planned routes better.'
What did you build yourself?
Many projects are team work. She wants to know your own part clearly.
Use 'I' for your own work and 'we' for the team. Name two or three things you did yourself, such as cleaning the data, choosing features or testing the model. Be honest about what others did.
What changed after the model was in use?
She wants to see that your work helped people make better decisions.
Give one result. It can be a number, but it does not need to be. 'The support team answered urgent tickets first' is a good answer. If you do not know the exact number, say so honestly.
Why did you choose that method?
She wants to understand how you think, not just which tools you know.
Give a clear reason, such as accuracy, speed, or how easy it is to explain. Do not only list tools or buzzwords. Say what the problem needed and how your method fit it.
What else did you try, and why did you not pick it?
Naming another option shows you compared choices carefully.
Name one other method and one trade-off. For example: 'A more complex model was a little more accurate, but it was harder to explain. The simple model was nearly as good and the team trusted it.'
How did you weigh accuracy against speed or simplicity?
Real work always has trade-offs. She wants to see good judgement.
Explain what mattered most for the users. Say who used the results and what they needed. Use plain words, such as 'Speed was more important here, because the result was needed in seconds.'
How do you work with a product team?
Your models are only useful if the product team can act on them.
Give one concrete example. Say how often you met, what you shared and what you asked. For example: 'Each week I showed the product manager one chart and asked what decision it helped.'
How did you agree on what to measure?
Good goals at the start save time later. She wants to see you plan with others.
Describe one meeting or conversation. Say who was there, what goal you agreed, and why that measure made sense for the product. Keep it short and real.
How did you explain your results to the product manager?
She wants to know if non-technical people can understand and use your work.
Show that you used plain words and simple charts. Give one example sentence you used. Say what the product manager decided after your explanation.
Tell me about a time a model did not work as you hoped.
She wants to see honesty and how you learn from problems.
Choose a real, small problem. Say what went wrong, what you did next and what you learned. Stay calm and positive. Do not blame other people.
Let me start with the problem the model solved.
Clara
AI coach
Leo
AI coach
Lines to try
Let me start with the problem the model solved.
My own part was cleaning the data and choosing the features.
The other option was more accurate, but harder to explain.
We chose a simpler model because the team needed to trust it.
I shared the results with the product team in plain words.
Could you say that again, please? I want to answer it well.
Common questions
What if my model only made a small change?
That is fine. Dana wants a real example, not a big number. Say clearly what changed, even if it is small. For example, 'The team planned routes better.' A small, honest result is better than a big, unclear one.
Should I use a lot of technical words?
Use only the words you need. Dana cares about clear thinking more than clever words. When you name a method, always give a reason. Practise explaining your work as if you are talking to a product manager.
What if I do not understand a question?
It is normal to ask. You can say, 'Could you say that again, please?' or 'Do you mean the model or the project?' Taking a moment to think shows care, not weakness.