Zycus Machine Learning Engineer Interview Experience
Zycus is one of the pioneers in cognitive procurement. I applied in the Machine Learning Engineer role from LinkedIn.
In the first round they asked some questions like:
- Total IT experience
2. Relevant Experience in Machine Learning
3. Relevant experience in NLP-
4. Have you worked on Seq2seq model?
5. What is your understanding on anomaly detection, quantity predictions models and algorithms?
6. Relevant experience in Deep Learning?
7. What is your understanding on LSTM Architecture/LSTM classification?
8. How is LSTM used in Text classification?
9. Which deep learning frameworks have you used?
10. Have you worked on multilabel classification model using bidirectional lstm?
11. Have you worked on recommendation engine?
12. Have you worked on Name entity recognition or built any model using NER?
13. Have you built Entity extraction model from large corpus of unstructured data?
14. Any experience or Knowledge on OCR.
15. Qualification and university passed out from.
16. Notice Period
17. Current CTC
18. Expected CTC
I got shortlisted in the first round. The Second Round was a telephonic interview round with the director of the company. He gave me one usecase
on machine learning. It was quite a long question.
The third round was an assignment round. The assignment was to build an AI/ML model for extracting agreement date, agreement value, start date, end date, party one and party two from various rental agreements. Training set contained around 30 doc files and test set around 7.
The fourth round was direct video conferencing round of 1 hour with the directors. They asked questions on
1) How to do feature engineering for categorical features and numerical features?
2) How to deal with imbalanced dataset?
3) How to deal with outliers?
4) how to create a model with chest X-Ray images?
5) one use case regarding translation invariance.
6) if there is image data and we add some numerical features how will you deal with it?
7) Remaining was regarding my projects from CV.
8) They also asked some more use cases I can’t remember properly.
