Candidatei-me por indicação de um funcionário. O processo levou 3 semanas. Fui entrevistado pela Observe.AI (Bengaluru) em jun. de 2020
Entrevista
There were several rounds.
1. Discussion with the Hiring Manager
2. Technical Round (Basics)
3. Technical Round (Projects and experience)
4. Technical Round (Hands-on Round)
5. Another discussion with the hiring manager.
6. Discussion with HR.
All the interviews were done through video calls.
The whole process took around 3 weeks.
Talent acquisition team was very prompt with replies for all my queries.
I was sent a detail about what the next discussion will be focusing on beforehand. This helped in my preparation.
I personally enjoyed the Hands-on round. The interviewer helped me understand the problem and guided me to arrive at a solution through the discussion. The questions were unique and intuitive.
Over all it was a positive experience. All the interviewers were more focusing on Knowledge that we posses and the depth of it.
Perguntas de entrevista [1]
Pergunta 1
Basic questions from : Bias and Vairance trade-off, CNN, SVM with a little twist. Decision Tree, Random Forrest. Pros and Cons of these techniques. Detailed discussion on any Deep learning Architecture. ex: Transformers, wav2letter++ etc.
Candidatei-me por indicação de um funcionário. Fui entrevistado pela Observe.AI (Bengaluru) em abr. de 2021
Entrevista
The interview consisted of 5 rounds:
1. Initial ML architecture round wherein the interviewer asked about my projects, the algorithms used, the theory behind them etc.
2. Hands on ML coding round: Understanding ML algorithms and coding them up.
3. ML architecture round: Designing ML algorithms in detail. Attention to detail is the most important.
4. Hiring manager round: Project discussion and explanation on architectural decisions
5. HR discussion
Perguntas de entrevista [1]
Pergunta 1
Popular ML components like LSTMs, Transformers, CNN etc. Building ML models in the context of speech recognition/NLP and advantages or shortcomings of each approach. These have to all be addressed while building an ML pipeline. Know your own projects in detail and why certain models were used vs. others
