Capgemini

There were two technical rounds. It was an easy process. Questions mostly covered Generative AI core concepts, Python coding, basic machine learning questions, and about projects. Overall, it took two weeks.

Basic GenAI questions, including architecture-level topics such as how transformers work, the difference between RNN and LSTM, why transformers are used instead of RNNs, and their drawbacks. This also includes some use case development.

It was more about problem-solving and the usage of new LLM models: RAG, Transformers, Ollama, and quantization. Faced two rounds of interviews but was rejected in the second round due to a lack of deployment experience.