Me postulé a través de un reclutador. Acudí a una entrevista en IBM en oct 2024
Entrevista
Data Science questions, ML questions, and some Python programming questions are easy to medium in difficulty. They cover overfitting, how a Random Forest model works, the difference between bagging and boosting, the transformer architecture, and some Pytorch-related questions.
Preguntas de entrevista [1]
Pregunta 1
Q: what is the difference between Pytorch and Tensorflow (computational-wise)
Q: describe the transformer architecture
Q: difference between encoder-based models and decoder-based models
Me postulé en línea. Acudí a una entrevista en IBM (Estrasburgo)
Entrevista
1er entretien RH assez concis avec une présentation rapide de mon parcours
2nd entretien technique avec Manager des équipes data de Lille et Strasbourg : présentation de mon parcours puis question sur une problématique donnée (liée au domaine bancaire)
3eme entretien technique ; présentation et question autour de l'IA générative
Preguntas de entrevista [1]
Pregunta 1
Quelles sont les métriques pour jauger les performances d'un LLM ?
Me postulé en línea. Acudí a una entrevista en IBM
Entrevista
"The company provided a fixed interview slot without any room for flexibility. As soon as I responded to ask if we could reschedule for a different date and time, the system automatically triggered a rejection email. It seems their automated process drops candidates who cannot make the initial slot.
Quiet straight forward, but unfortunately got rejected even afer passing all unit tests with same old saying that other candidates are better aligned at the moment for this specific role demands.