Persistent Systems AI Engineer Interview Questions

Fit a univariate linear regression model from data using gradient descent or the normal equation.

MathArraysGradient DescentPersistent Systems

Explain how to select metrics, validate predictions, and analyze errors when evaluating a machine learning model.

model performanceevaluation metricsPrecisionPersistent Systems

Implement a decision tree classifier from scratch, including split selection, stopping rules, prediction, and evaluation.

Feature Engineeringmodel trainingModel EvaluationPersistent Systems

Key pipeline considerations for deploying an ML model into production, including orchestration, reproducibility, data quality, and monitoring.

InfrastructuremonitoringQualityPersistent Systems

Build a reusable NLP preprocessing pipeline with normalization, tokenization, stopword filtering, and lemmatization.

LemmatizationStemmingNLPPersistent Systems

Explain how to evaluate RAG retrieval, generation quality, factual grounding, latency, and production impact.

factual groundingfeedback loopfailure modesPersistent Systems