Tiger Analytics GenAI Engineer Interview Questions

Design a leakage-safe text preprocessing pipeline that removes noise while preserving information needed for transformer fine-tuning.

data preprocessingdata cleaningDeep LearningTiger Analytics

Compare production LLM deployment architectures and explain trade-offs across latency, cost, quality, reliability, and operations.

llm deploymentproduction systemsarchitectureTiger Analytics

Design a production deployment path for a personalized ranking model, with serving, feature consistency, drift handling, and experiment driven rollout.

InfrastructureFeature DriftModel ServingTiger Analytics

Discuss how you designed an LLM system for a business use case, including evaluation, hallucination control, and cost latency tradeoffs.

Structured ExtractionPrompt EngineeringLLM EvaluationTiger Analytics

Explain top-k versus top-p sampling and select the right decoding strategy for controlled, diverse LLM outputs.

llm basicsinference optimizationmodel comparisonTiger Analytics

Improve SQL query performance on large datasets using execution plans, indexing, partitioning, and efficient ETL patterns.

ETL optimizationInfrastructureindexingTiger Analytics