Why OnePay? How do you understand frauds for any fintech company? Design a ML System for fraud detection?
Applied Scientist Interview Questions
1,159 applied scientist interview questions shared by candidates
Design an experiment for *insert business case*.
basic probability and coding question in the screening
They ask me about how to design a metric to measure the time to which the food is prepared vs. the rider is arrived at the shop.
How to match drivers and riders in an airport
Describe what you have done in your current role as a data scientist.
Coding question: search for an element in a sorted matrix
Machine Learning Theory Interview (By Hiring Manager) 1. Interviewer started asking about my thesis and asked questions about IMU sensor, Odometry 2. Classification metrics (Precision, Recall, curves - what is measured along X, Y and how is area calculated, how you conclude about best threshold) 3. Decision Trees and Random forest regularization techniques 4. Advantages of LLM over LSTM 5. Explain LLM architecture and every layer functionalities 6. Scenario-based question: How would you ensure a particular class is distinguished correctly during training and achieves zero errors? (By assigning a higher weight to that class during training.) 7. Explain about Adaboost and how you can modify according to your dataset 8. Scenario based: If the Product Manager provides a new requirement, how would you approach and execute it? 9. Scenario based: Explain a scenario from your prev experience where you had to work with multiple stakeholders 10. Scenario based: You are given a requirement to develop a solution that relies on external AI APIs. However, the Product Manager does not want any sensitive data to be sent outside the system. How would you handle this situation? I would initially use the external APIs but ensure that all sensitive user data is masked or anonymized before sending any requests. This allows rapid prototyping while keeping data privacy intact. In parallel, I would begin working on developing an in-house model capable of handling the required tasks, so that we can eventually migrate away from external APIs entirely. (This was the same approach eventually adopted by the team.)
Top Grading Discussion (By Director - Applied Science) This round involved a deep dive into my projects, my role at Walmart, and my hackathon experiences. The interviewer focused on understanding my overall approach, problem-solving process, and decision-making rationale. I was asked questions such as why I selected particular models, what training strategies were used for fine-tuning, how knowledge distillation was applied, and other aspects that revealed technical depth and reasoning.
deep learning and machine learning related stuff
Viewing 981 - 990 interview questions
See Interview Questions for Similar Jobs
Deep Learning ScientistMl EngineerData Scientist NlpSenior Machine Learning ScientistDeep Learning Data ScientistApplied Research ScientistNlp Data ScientistAi Machine Learning EngineerDeep Learning Research EngineerApplied ResearchJr Data ScientistArtificial Intelligence ScientistData Scientist TraineeDeep Learning / Machine Learning ResearcherJunior Machine Learning EngineerApplied Machine Learning EngineerSenior Machine Learning EngineerMachine Learning Researcher