Screening Background, projects, motivation. Basic Python (lists, dicts, file handling). Quick debugging task. Data Foundations Pandas: cleaning, groupby, missing data. Matplotlib: plots, subplots, customization. Hands-on: load → process → visualize dataset. ML & PyTorch PyTorch basics: tensors, autograd, training loop. Build & explain a simple neural net. Optimizers, overfitting, regularization. Transformers (75 mins) Attention mechanism & architecture. Hugging Face: tokenizers, pretrained models. Case study: fine-tune BERT for classification. RAG Design Retriever + generator concept. Pipeline with embeddings & vector DB. Use case: Q&A over company docs. Discuss scalability, hallucinations, retrieval quality. Final Round End-to-end design: RAG assistant for financial documents. Covers preprocessing, embeddings, retriever, fine-tuning, evaluation. Behavioral: teamwork, ownership, learning mindset.

It was for associate ML engineer position which was 1-hour technical round with several machine learning questions. There were two simple Python problem-solving questions. The interview included numerous theoretical questions on machine learning.

CV screening, Followed by Technical review and HR review. Then salary negotiations. Then orientation by the head of department. SIx months probation periods. Then only full time employee. Then only work work no other options than work