Tech Abstractions

Practice Questions

ML system design interview questions with AI-powered scoring.

Design a Context Management System for LLM AgentsMediumDesign a Multi-Agent Orchestration SystemMediumDesign a Simple Conversational Agent for Customer SupportEasyDesign Agent Memory Architecture for Long-Running TasksHardDesign an Agent Evaluation and Monitoring FrameworkMediumDesign Tool-Use Safety Mechanisms for Autonomous AgentsHardDesign a Feature Store for a Machine Learning PlatformHardDesign a Fraud Detection System for FintechMediumDesign a Multi-Tenant ML Inference PlatformMediumDesign a Real-Time Recommendation System for E-CommerceHardDesign a Search Ranking SystemEasyDesign an Ad Click Prediction Pipeline at ScaleHardChoose the Right Autonomy Level for a Financial Anomaly SystemMediumDebug a Fraud Model with Segment-Level DegradationHardDefend a Proxy Label Choice Under VP Cross-ExaminationMediumDefend Three Months of Problem Framing to a Skeptical VPEasyDesign a Continuous Training PipelineHardDesign a Data Validation Pipeline for MLEasyDesign a Model Deployment and Rollback StrategyMediumDesign a Model Monitoring and Alerting SystemHardDesign an ML Experiment Tracking PlatformMediumDesign an ML Feature Engineering PlatformEasyDetect a Silent Framing Failure When Model Metrics Look FineMediumDiagnose a Churn Model That Passed Eval but Moved No KPIHardFrame a Home-Page Recommendation System End to EndHardFrame a Real-Time Content Moderation System Under Hard ConstraintsHardFrame a Vague ML Request Using the Four-Layer StackEasyHandle Multi-Week Label Lag in a Search Ranking ReframeMediumInherit and Fix a Broken Recommendation SystemMediumReframe a Multiclass Classifier That Cannot ScaleMediumRules vs. ML for Support Ticket RoutingEasyScore a Proxy Label on Five DimensionsMedium