NMH Interview Questions
The questions to prepare for NMH interviews, across all roles. Questions from real interview reports rank first. Updated daily.
Design an LLM serving architecture that manages long generation latency, GPU efficiency, reliability, and model quality.
Design a low-latency, cost-aware serving platform for multiple fine-tuned LLMs under variable traffic.
Compute cosine similarity for equal-length high-dimensional vectors using dot products and Euclidean norms.
Design a practical framework for comparing LLM quality, reliability, latency, cost, and safety in production.
Design a token-aware, hierarchical document processing pipeline that preserves global context despite language-model context limits.
Explain a production RAG pipeline and techniques for reducing unsupported or fabricated answers.
Compare embedding models for domain-specific semantic search using retrieval quality, latency, memory, and domain adaptation.
Design an end-to-end RAG pipeline with transformers, grounded generation, evaluation, and justified fine-tuning decisions.
Sign up to see every question
Create a free account to unlock this list and practice real interview questions.