Tell us about your relevant experience with AI: both for accelerating development and as part of business solutions.
Machine Learning / AI
Was the data preprocessing manual or automated? Were models used in the pipeline?
Was there experience of fully delegating AI development to an agent without code review, and how did you systematically prevent issues like breaking neighboring components?
What AI tools and models do you use for personal and work purposes and why?
What problems did you encounter during the development of the RAG solution and what solutions improved the output?
Have you had any personal or internal MVPs outside of main work tasks using AI?
Do you have any other interviews at the final stage?
How did you prepare data before loading it into the vector database?
1. What will be the order of output in the console? Approximately how long will the entire program take to execute? 2. How to modify main() so that func1() and func2() run asynchronously, and the total execution time is reduced to about 2 seconds?
Are you currently continuing to work?
Briefly explain the mechanism of self-attention.
In which parts of Transformer LM is most of the weights concentrated and what is MoE used for?
Have you worked with Airflow?
Did you modify the prompt for the end user? What problems did the prompts solve?
Tell us a little about yourself and your experience.
How to bring an unstable TTS pipeline to industrial standard?
Model A has WER 3%, model B has WER 2%. Which one is better?
Can I just cut out all silence parts from audio?
Why did you decide to discuss a role that is more focused on analytics than Data Science, and what do you want to find in the next position?
What do you use to cut long audio files into speech segments?