Fresh Gravity Interview Questions
& Process

This was a 2 round interview. The first round had theoretical GenAI questions about transformers and RAG architecture. The second round covered the math behind transformers and other algorithms, ML models, LLM fine tuning, open source models, and basic statistics.

The interview process typically consists of four rounds. It starts with a coding round, followed by two technical interviews, each lasting an hour. The coding platform provided a subpar experience, with a code editor resembling a basic notepad that lacked the functionality to run code before submission. Candidates had only five submission attempts, and many of these failed due to unspecified 'file errors'. Some coding problems were also poorly formulated, such as one that required writing code without a main method and specifying an array size, while also prohibiting the use of collections. Given the company's long history in the market, a more professional setup was expected. I suggest the company adopt a reputable coding platform with proper compilers to ensure a fairer and more efficient candidate experience.

The interview process involved some easy Python coding questions, Python theory questions, and scenario-based questions about webhooks and Kafka. They also asked about my past projects. Overall, the questions were not too difficult, around an average level.

It was 3 rounds total. One technical, one techno-managerial, and a final HR round. Honestly, it felt like a waste of time. I already told them my expected salary, and HR said they had no budget issues. But in the final round, they said my expectation was way beyond their budget. They should be upfront about their pay range to avoid wasting candidates' time. If you can't meet expectations, don't conduct interviews.